Experimental calibration of a numerical model describing the alteration of cement/caprock interfaces by carbonated brine

Deep wells provide a possible pathway for CO2 and brine leakage from geologic storage reservoirs to shallow groundwater resources and the atmosphere. The integrity of wellbore cement in these environments is of particular concern, because it is not known if changes in cement properties resulting from reaction with CO2-rich brines will lead to enhanced leakage over the life cycle of the storage reservoir. Assessment of wellbore leakage will ultimately be answered through models that capture both the chemical and physical processes and the uncertainty of key parameters within the wellbore environment. Towards this end, we use the results for 13 core-flood experiments conducted at variable partial pressures of CO2, flow rate, durations, and cement-caprock apertures to constrain a wellbore model that couples chemical processes important to assessing the long-term integrity of wellbore cements in geologic carbon storage environments. X-ray computed microtomography collected prior-to and following the experiments was employed to spatially resolve the interface and the extent of the reaction zones, and time dependent solution chemistry was used to track the chemical alteration over the course of the experiments. In this manuscript we focus on the development of geochemical model that describes the alteration of both the cement and the caprock. In our experiments, chemical alteration of the cement significantly exceeded any dissolution of carbonate minerals within the caprock and fracture geometry played no role on the extent of reaction. The experimental data was used to calibrate a numerical model of wellbore-caprock interfaces coupling reaction-front chemistry, fluid flow and transport of dissolved species. The geochemical model adopts an idealized representation of the cement chemistry in which appropriate equilibrium conditions are enforced at a series of discrete reaction fronts. The equilibrium conditions are coupled by diffusive transport between the fronts, which also determines the rate of front propagation. Despite its simplicity, the calibrated model accurately reproduces the reaction-zone growth and effluent chemistry for the range of experimental conditions considered and allowed key parameters to be confirmed or calibrated. These include the use of portlandite, calcite, and analcime solubility as equilibrium controls at specific reaction fronts within the cement; the use of constant effective diffusivity for each alteration zone; and diffusive growth of the alteration layers. (C) 2014 Elsevier Ltd. All rights reserved.