Modeling Metal Transport After CO2 Leakage Into Aquifers From Carbon Sequestration: Relevance of Desorption and Mineral-Dissolution Kinetics

Carbon capture and storage is an important part of the collection of emerging technologies proposed to reduce carbon emissions.  Prior to widespread application of geologic carbon sequestration, it is important to understand human health risks in the event that carbon dioxide leaks into overlying drinking water aquifers.   One area of concern is the mobilization of metals as aquifer pH is reduced due to an increase in partial pressure of CO₂.  Desorption and dissolution are important metal mobilization mechanisms.  However, for many trace metals of concern, such as lead, arsenic, barium and nickel, literature values for desorption and dissolution rate constants vary over orders of magnitude. This variability is partially a result of the dependence of rate constants on pH, concentration, temperature, ion competition, presence of organic matter, aquifer mineralogy, and redox conditions.

Many reactive transport studies assume that desorption occurs instantaneously and therefore can be assumed to behave according to thermodynamic equilibrium.  The focus of this work is to consider metal and mineral combinations, as well as flow conditions, under which this assumption may not be valid.  An analysis of Damkohler numbers for available desorption rate constants over length scales of interest demonstrates that in the realm of typical groundwater velocities, metal sorption kinetics should be considered on length scales of approximately 1-100 meters.  In contrast, at length scales greater than 100 meters, metal mobilization can be modeled simply using linear sorption equations (i.e. Kd), however kinetic dissolution of slow dissolving minerals should be considered over all length scales relevant to groundwater modeling, unless the contribution from dissolution is negligible. Based on these results, the relative importance of desorption and dissolution kinetics are modeled in a one-dimensional flow model using PHREEQC.  Ultimately, this analysis will provide needed constraints on metal sources in sophisticated risk assessments.