Abstract

As CCUS deployment accelerates, repurposing existing oil and gas pipelines, wells, and subsurface infrastructure offers major economic and strategic advantages. However, corrosion under dense-phase and supercritical CO₂ (s-CO₂) conditions remains a critical integrity challenge, particularly when water, chloride, and reactive impurities are present. Although dry s-CO₂ is generally benign to steels, wet s-CO₂-rich phases and s-CO₂-saturated aqueous/brine environments can promote general corrosion, localized attack, and unstable corrosion-product layers. The large scatter in reported corrosion data reflects the combined effects of phase state, water availability, impurity chemistry, temperature, pressure, exposure time, alloy composition, and corrosion-layer evolution.

This talk presents our recent work on s-CO₂ corrosion from a mechanism-informed engineering perspective. It first summarizes key corrosion risk scenarios relevant to CO₂ transport, injection, storage, and EOR, highlighting why laboratory results must be interpreted within realistic operating envelopes; then discusses our experimental studies on candidate steels and alloys in s-CO₂-saturated saline environments, showing that FeCO₃ commonly forms as an outer corrosion product, while Cr-enriched inner layers can improve protection. In particular, alloys containing approximately ≥9 wt.% Cr tend to develop more compact and adherent corrosion layers, providing guidance for material selection. Beyond bulk Cr content, we introduce a transient amorphous layer framework for early-stage corrosion, showing how Fe–C-, Fe–O-, and Cr–O-dominated interfacial layers regulate CO₂-derived adsorption and reaction pathways, helping explain non-monotonic corrosion trends. Finally, we discuss how curated datasets and machine-learning models can support corrosion risk mapping, impurity prioritization, and safer reuse of existing infrastructure for CCUS.