Global Sensitivity Analysis of Reactive Transport Modeling of CO2 Geological Storage in a Saline Aquifer

Abstract

Predictive uncertainty for reactive transport modeling of CO2 geological storage arises due to high uncertainty in dissolution/precipitation rates. Here, the reactive transport modeling of the Frio sandstone formation is used as a case study. The major CO2 trapping mineral is ankerite, while the main dissolution minerals are oligoclase and chlorite. In this context, unlike the commonly used local sensitivity analysis, the sensitivity analysis is global so that the potential co-operative effects among input parameters can be investigated. Nine key factors for kinetic rates and reactive mineral surface areas with respect to precipitating and dissolving minerals (only oligoclase) are considered. Sensitivity results from the Morris method show that the dissolution rate of oligoclase knu_O, and its reactive surface area A_O, are the most sensitive parameters, with the largest effects on CO2 mineral capture. The variation of the total amount of CO2 captured by minerals is pronounced with multiple model runs from Morris samples, which suggests that reactive surface areas and kinetic rates have significant impacts on CO2 mineral sequestration.