Spatial Colorimetric Mapping of pH and Dissolved Carbon during Convective Dissolution of CO{\_2}: Constraints from Domain, Geometry, and Indicator Mix on Measurement Reliability

Quantifying the dissolution and buoyancy-driven transport of CO2-rich water is critical to understanding solubility trapping mechanism in geologic storage. Analog-fluid density driven convection experiments do not represent CO2 dissolution chemistry or the aqueous CO2 density-concentration monotonic relationship. X-ray tomography does not provide enough contrast to directly measure dissolved CO2. Spatial colorimetry provides an inexpensive method to map pH and estimate total dissolved inorganic carbon (CT) via carbonate equilibrium assumptions. However, the effects of experimental conditions such as domain (free fluid and porous), geometry (2D and 3D), and indicator mix concentration on the reliability of pH and CT measurements remain unexplored. We evaluated how these factors affect the reliability of measurements using two different colorimetric models (hue and spherical coordinates (θ and ϕ)). Hue was reliable in all domains whereas θ-ϕ was reliable in free fluid. Lower indicator mix concentration increased the chromatic sensitivity at low pH in free fluid; however, it did not produce sufficient color contrast for pH calibration and mapping in the porous domain. In the porous domain, θ-ϕ calibration was 33 times more susceptible to noise than hue. Temporal pH noise was less than 0.01 units; spatial noise changed domain-mean CT by 0.6-4.4{\%}. Both models were able to resolve CO2-rich plumes in free fluid; however, θ-ϕ had sharper plume boundaries than hue. The temporal evolution of CO2 dissolution resolved by the color models agreed more in 2D than 3D. These findings provide a reliability framework for studying subsurface dissolution and mixing of CO2.