11376 Effect of Debonded Interfaces on Corrosion of Mild Steel Composites in Supercritical CO2-Saturated Brines

Thursday, March 17, 2011: 8:25 AM
Room 352 A (George R. Brown Convention Center)
Jiabin Han*1, J. William Carey1, and Jinsuo Zhang2
(1)Earth and Environmental Sciences Division, Los Alamos National Laboratory; (2)Decision and Application Division, Los Alamos National Laboratory
The geologic sequestration of CO2 is a proposed method to limit greenhouse gas emissions and has been the subject of many studies in the last decade. The viability of CO2 sequestration depends on the long-term performance of wellbore systems that must prevent buoyant supercritical CO2 from escaping the storage reservoir. Wellbore systems achieve isolation of the storage reservoir through a combination of steel (generally carbon steel) and Portland cement. CO2 leakage path development in cement has been widely investigated.  However, field analyses showed that the interfaces along the cement-caprock and cement-steel interfaces were more significant potential pathways for CO2 leakage and are crucial to wellbore integrity. In particular, CO2 leakage along the steel-cement interface has the potential to accelerate corrosion. This issue has received very little attention, and in this paper, we conduct experiments to assess the corrosion risk at cement-steel interface under in situ wellbore conditions.

Wellbore interfaces were simulated by assemblies constructed of J55 mild steel and Portland class G cement and corrosion was investigated in supercritical CO2 saturated brines, (NaCl=1 wt%) at T=50 °C, pCO2=1500 psi with interface gap size = 20 um, 100 um and ∞ (open surface).  The experiments were carried out in a high-pressure, 1.8 L autoclave.  The corrosion kinetics were measured employing electrochemical techniques including linear polarization resistance, electrochemical impedance spectroscopy, and Mott-Schottky techniques.  The corrosion scales were analyzed using secondary electron microscopy, back scattering electron microscopy and energy dispersive spectroscopy.

For initially fresh surfaces, corrosion rates decreased as the interface gap decreased.  In this case corrosion rates are controlled by diffusion rates of corrosive species through the interface gap.  Passivated steel was created in one week experiments and confirmed by observations of increased corrosion potential and decreased corrosion rate.  Passivated steel corrosion rates were two orders of magnitude less compared with fresh steel.  The corrosion scale is amorphous for the open surface.  Well-crystallized scale was observed at interface gap sizes <100 µm.  All corrosion scales were composed of co-precipitated iron and calcium carbonates.  Mott-Schottky measurement showed that the corrosion scale had semiconductor properties, indicating that the corrosion scale must consist of additional phases because FeCO3 and CaCO3 are non-conductive.  An additional semi-conductive layer must exist and is believed to play a crucial role in retarding corrosion or passivating steel surface.  The property of this scale will be further explored in our ongoing research.