Wellbore integrity and CO2 -brine flow along the casing-cement microannulus

Abstract

Wellbore integrity is one of the key performance criteria in the geological storage of CO2. This is significant in any proposed storage site but may be critical to the suitability of depleted oil and gas reservoirs that may have 10’s to 1000’s of abandoned wells. Much previous work has focused on Portland cement which is the primary material used to seal (create zonal isolation) wellbore systems. This work has emphasized the reactivity of Portland cement to form calcium carbonate. However, an increasing number of field studies [e.g., 1], experimental studies [e.g., 2], and theoretical considerations indicate that the most significant leakage mechanism is likely to be flow of CO2 along the casing-cement microannulus, cement-cement fractures, or the cement-caprock interface. The magnitude of flows along these interfaces is a complex function of the pressure gradient, geomechanical properties that support the interface and dissolution/precipitation reactions that lead to widening or closure of the interface. In this study, we investigate the casing-cement microannulus through core-flood experiments. The experiments were conducted on a 5-cm diameter sample of cement that was cured with an embedded rectangular length of steel casing. Prior to the experiment, the casing was loosened creating a poorly bonded interface. However, we discovered that under confining pressure this interface was non-transmissive, suggesting that in the wellbore environment an open casing-cement microannulus requires a relatively low differential between pore and confining pressure. For the experiments, we created an artificially transmissive interface by scoring grooves in the steel casing (0.2-0.8 mm in depth). The core-flood experiments were conducted at 40 ∘C, 14 MPa pore pressure, and 28 MPa confining pressure for a period of 400 hours. During the experiment, 6.2 L of a 50:50 mixture of supercritical CO2 and a 30,000 ppm NaCl-rich brine flowed through 10-cm of limestone before flowing through the 6-cm length cement-casing composite. Approximately 41,000 pore-volumes of fluid moved through the casing-cement grooves. Scanning electron microscopy revealed that the CO2-brine mixture impacted both the casing and the cement. The Portland cement was carbonated to depths of 50–150 μm by a diffusion-dominated process. There was no evidence of mass loss or erosion of the Portland cement. By contrast, the steel casing reacted to form abundant precipitates of iron carbonate that lined the channels and in one case almost completed filled a channel. These results are compared to field studies to constrain the magnitude of possible CO2 migration in real wellbore systems.