Dual-stage framework for corrosion prediction in oilfield conditions combining short-term data and long-term extrapolation

Purpose This study aims to develop a reliable predictive framework for estimating long-term corrosion behavior and service life of pipeline materials exposed to high-pressure CO2 rich oilfield environments. Traditional models often fail to capture the dual-phase nature of corrosion – initial rapid degradation followed by stabilization – leading to inaccurate lifetime predictions. By integrating short-term experimental data with long-term extrapolation using power-law and exponential decay models, the proposed dual-stage framework addresses this gap and supports safer, more cost-effective operations in the oil and gas industry. Design/methodology/approach The framework was applied to L80-1, P110 SS and 2205 DSS materials exposed to simulated oilfield environments with CO2 partial pressures of 0.5, 1.1 and 2.6 MPa. Corrosion rates were measured at 7, 14, 21 and 28 days, and fitted using power-law and exponential decay models to extrapolate long-term behavior. Model performance was evaluated using statistical metrics (R², root mean square error and mean absolute error), and service life was estimated based on cumulative material loss compared to critical wall-thickness thresholds. Findings The dual-stage framework effectively predicts corrosion behavior and service life. The power-law model captures early corrosion dynamics, while the exponential decay model excels at modeling long-term stabilization, especially for corrosion-resistant materials. Results closely matched experimental data, with 2205 DSS showing the longest projected service life, followed by P110 SS and L80-1. Originality/value This study presents a novel, validated approach that improves prediction accuracy by accounting for both corrosion phases. It aids in material selection and maintenance planning, offering practical value to researchers and oilfield operators.