Predicting impacts of CO2 intrusion into a confined sandstone aquifer

Deep subsurface storage and sequestration of CO2 has been identified as a potential mitigation technique for rising atmospheric CO2 concentrations. Sequestered CO2 represents a potential risk to overlying aquifers if the CO2 leaks from the deep storage reservoir. Experimental and modeling work is required to evaluate potential risks to groundwater quality and develop a systematic understanding on how CO2 leakage may cause important changes in aquifer chemistry and mineralogy by promoting dissolution/precipitation, adsorption/desorption, and redox reactions. Sediments from a confined sandstone aquifer, i.e., the High Plains aquifer in Kansas, were used to represent a generic sandstone aquifer. The sediments originated from different wells and depths within the central portion of the High Plains aquifer. A series of batch and column experiments were conducted to study time-dependent release of major, minor and trace elements when the sediments were exposed to the CO2 gas stream. Pre- and post-treatment solid phase characterization studies and wet chemical extractions have also been conducted or are underway. Major variables tested included reaction time (0-336 hours), CO2 flow rate (50 to 350 ml/min), brine concentration (0.1 and 1 M NaCl), and sediment type. Additional experiments are being conducted to determine the fate of contaminants, such as As, Pb and Cd, when they are present in the initial contacting solution. The XRD results showed that the < 2 mm size-fraction of the High Plains aquifer sediments was abundant in quartz and feldspars, and also contained 15 to 20 wt% montmorillonite and up to 5 wt% micas. Some sediments contained up to 7 wt% calcite. Results from acid extractions demonstrated that the solid phase had appreciable amounts of potential contaminants (As, Cd, Cu, Pb and Zn). However, results from the batch and column experiments demonstrated that few trace elements were released into the aqueous phase and their concentrations were close to or below detection limits. The concentrations of other elements, such as Si, Ca, Ba, Mg, Mn, Sr, Na and K increased either instantaneously or followed nonlinear increasing trends with time, indicating that they were controlled by dissolution and/or desorption reactions. Reactive transport models were developed to interpret the concentration changes observed in experiments conducted with the High Plains aquifer sediments. The initial conceptual model was developed based on literature data collected from other sites and tests. Initial assumptions were that the release of major and trace element were caused mainly by the calcite dissolution and Ca-driven cation exchange reactions. The necessary changes made in the initial conceptual model reflect the site-specific nature of the impact of the leaking CO2 on the groundwater quality. The results from these investigations will provide useful information to support site selection, risk assessment, and public education efforts associated with geological CO2 storage and sequestration.