Fracture susceptibility and depressurization dynamics in CO2 pipelines under uncertainty

Since the success of large-scale carbon capture and storage (CCS) infrastructures relies on the safety of pressurized CO 2 pipelines, quantifying the susceptibility to running ductile fracture during planned or accidental depressurization is still a key challenge in the field. In this work, we present an integrated framework for uncertainty quantification in CO 2 pipelines that combines a high-fidelity, one-dimensional transient solver based on the Homogeneous Equilibrium Model with a reduced-order surrogate constructed via Polynomial Chaos Expansion. The methodology is applied to two representative infrastructures, the Cortez and Weyburn pipelines, to analyze fracture-related metrics based on the Battelle Two-Curves Method and the depressurization dynamics under uncertainty in operating conditions typical of CO 2 transport. Our analysis reveals that the pipe diameter (and its thickness) and the initial temperature are the main parameters in determining the pipe vulnerability to the running ductile fracture rather than the initial pressure. Specifically, only if the initial temperature is below a certain value (that depends on the pipeline), the onset of the fracture is prevented. In addition, we show that the minimal temperature reached during the depressurization has a similar behavior, while the discharged mass, relevant in the case of leaks, depends on both the initial temperature and pressure. Overall, the proposed approach achieves high predictive accuracy while drastically reducing computational cost, providing a basis for risk assessment to support safer and more robust CCS pipeline design and operation. • We propose an integrated framework for uncertainty quantification in CO 2 pipelines. • High-fidelity simulations based on the transient homogeneous equilibrium model. • The reduced order surrogate is constructed using Polynomial Chaos Expansion. • The temperature is the dominant parameter for predicting fracture susceptibility. • The framework achieves high predictive accuracy while reducing computational costs.