Large scale deployment of CO2 geological sequestration requires the assessment of the risks. One of the potential risks is the impact of CO2 leakage on shallow groundwater overlying the sequestration site.The understanding of the key chemical processes and parameters are critical for building numerical models for risk assessment. Model interpretation of laboratory and field tests is an effective way to enhance such understanding. As part of this investigation, column experiments in which the CO2 saturated synthetic groundwater flowed through a column packed with materials from the High Plains aquifer, were conducted. Changes in concentrations of several constituents in the column effluent and pH were determined. In this paper, a reactive transport model was developed to describe and interpret the observed concentration changes, attempting to shed light on the chemical reactions and mechanisms and key parameters that control the changes in effluent chemistry. The reactive transport model described fairly well the changes in pH and the concentration changes of Ca, Mg, Ba, Sr, Cs, As and Pb. Calcite dissolution and Ca-driven cation exchange reactions were the major drivers for the concentration changes of Ca, Ba, Sr, and Cs. The pH-driven adsorption/desorption reactions led to a concentration increase of As and Pb. The volume fraction and reactive surface area of calcite, CEC and sorption capacity were key parameters in controlling the magnitude of concentration increase. Model results also showed that Ba, which is an important chemical element released into the aqueous phase during these experiments, may be incorporated into the calcite crystal structure and the dissolution of Ba-bearing calcite could be an alternative pathway to explain the increase in aqueous Ba concentration when sediments are exposed to the CO2 saturated leaching groundwater.

The National Risk Assessment Partnership (NRAP) has developed a suite of tools to assess and manage risk at CO2 sequestration sites. The NRAP tool suite includes the Aquifer Impact Model (AIM), which evaluates the potential for groundwater impacts from leaks of CO2 and brine through abandoned wellbores. There are two aquifer reduced-order models (ROMs) included with the AIM tool, a confined alluvium aquifer, and an unconfined carbonate aquifer. The models accept aquifer parameters as a range of variable inputs so they may have broad applicability. The generic aquifer models may be used at the early stages of site selection, when site-specific data is not available. Guidelines have been developed for determining when the generic ROMs might be applicable to a new site. This paper considers the application of the AIM to predicting the impact of CO2 or brine leakage were it to occur at the Illinois Basin Decatur Project (IBDP). Results of the model sensitivity analysis can help guide characterization efforts; the hydraulic parameters and leakage source term magnitude are more sensitive than clay fraction or cation exchange capacity. Sand permeability was the only hydraulic parameter measured at the IBDP site. More information on the other hydraulic parameters could reduce uncertainty in risk estimates. Some non-adjustable parameters are significantly different for the ROM than for the observations at the IBDP site. The generic ROMs could be made more useful to a wider range of sites if the initial conditions and no-impact threshold values were adjustable parameters. © 2017 Society of Chemical Industry and John Wiley & Sons, Ltd.