Probabilistic Data Envelopment Analysis Applied to Risk-Based Inspection of Flexible Pipes

Flexible pipelines and risers are widely used by oil and gas operators in Brazil in their subsea systems for offshore operations. These operators have teams of multidisciplinary specialists working on the integrity management of their assets, as well as a large amount of inspection data, a database of real failures, and failure predictive tools. However, inspection scope planning is based on the judgment of the experts operating the assets. With the advent of new technologies, the use of data science and decision-making tools has helped engineers to make more assertive and less subjective decisions. To ensure the operational safety of flexible pipes and risers, identifying the main failure modes that can lead to the total loss of structural integrity and all the failure mechanisms that will lead to each failure mode is essential. This work will present a risk-based inspection methodology for offshore flexible pipes and risers based on data envelopment analysis (DEA). The present methodology considers the organization of each failure mechanism through event trees, allowing the visualization of the chronological occurrence of non-conformities in stages until the failure mode is reached. Recent work has shown risk analyses using data envelopment analysis based only on events (nonconformities) detectable by inspection. The proposed methodology establishes a relative risk ranking for a group of flexible pipes based on event trees, formed by events detectable or not during inspection, considering their probability of occurrence over time.