The long-term corrosion behavior of abandoned wells under CO2 geological storage conditions: (1) Experimental results for cement alteration

摘要

For a feasible risk assessment of CO2-leakage through wells of the carbon dioxide storage sites at abandoned wells of aquifers and depleted gas fields, the CO2 reactivity and permeability of casing steel, pseudo formation and cement-plug need to be evaluated experimentally as functions of temperature, pressure and formation water chemistry to provide fundamental information for the assessment. Two types of laboratory experiments were conducted. These experiments are shown below: (1) Conventional batch-reaction experiments of cement cores in the system of CO2 and simulated formation water as 0.5 M NaCl, at 50 and 70 °C, 5, 8, 18 MPa for 100-1600hr. In this experiment, individual cement cores are allowed to react with wet CO2 and NaCl solution charged in titanium reaction vessels with PTFE separators. (2) CO2-injection reaction involving casing (API Grade J-55)-cement (API class A) and cement (API class A)-shale composites which were saturated with 0.5 M NaCl solution. As a counter experiment for batch experiments, zero CO2 runs for these cements were additionally carried out at 50 and 70 °C at 5 MPa for 100 and 400 hr. The CO2-injection runs were carried out at 50 °C and 8.5 MPa with a constant differential pressure of 5 kPa. The resultant products were used to analyze alteration depth via micro-focused X-ray computed tomography (μ-XCT) and electron probe micro analyzer (EPMA). For observation and determination of alteration phases, field mission scanning electron microscopy with energy dispersive spectroscopy (FESEM-EDS) and micro X-ray diffractometry (μ-XRD) were also performed. The alteration zones identified in the both cement cores of A and G in wet CO2 showed spatial developments of zones appearing in μ-XCT images as a function of square root of time (t~(1/2)), which can be interpreted as a diffusion-limited reaction. However, in the NaCl solution, these cement cores developed little alteration zones and poorly displayed time-dependency after 100 h. In the reaction system with wet CO2, alteration directly proceeds carbonation from portlandite (Ca(OH)2) and calcium-silicate-hydrate (C-S-H: e.g., similar to tobermorite: Ca5Si6O_(16)(OH)2·4H2O) into carbonate such as calcite (CaCO3). While, in the system with CO2-saturated NaCl solution, aqueous alteration can evolve ettringite (Ca6Al2(SO4)3(OH)_(12)·26H2O) as an essential compound in the cements into Friedel's salt (Ca4Al2O6Cl2·10H2O), as well as or prior to carbonation. Also generally in the aqueous system, the CO2 transport into cement is much more limited than that in the CO2 gaseous system. It is possible that the cement degradation by carbonation is inhibited in the NaCl solution system by this different alteration mechanism. In the CO2-injection runs, both composites of casing-cement and cement-shale allowed CO2 to pass earlier (breakthrough within several hours) but stopped CO2-flow finally for at least 11 hrs. The μXCT and EPMA investigations recognized that the dominant flow path of CO2 could be a micro-annulus, and the path could have been closed by precipitating altered phases such as carbonate and Na-K rich fine silicate residues. All of these results suggest that wet and saline condition may preserve the cement-plug from subsequent CO2 attack at the CO2 storage sites, even if the casing has micro-annulus. Although the cementitious materials at the deep CCS sites can be replaced with carbonates, the rate of replacement can be strongly reduced while the system maintains wet condition.