Multimodal Risk Evaluation Approach for Autonomous CO2 Capture Subsea Systems

Abstract Current case study presents state-of-the-art multimodal structural risks analysis approach, that is particularly suitable for multidimensional energy/transport systems, providing an alternative to existing univariate/bivariate (1D/2D) reliability methods. Since high-dimensional systems possess nonlinear intercorrelations between their principal components or dimensions, existing reliability methods that deal with dynamic time series struggle to handle structural system's high-dimensionality. Expansion of generalized extreme value reliability and statistics from 1D (univariate) toward 2D (bivariate) case meets with practical obstacles. First, extreme value theory (EVT) being univariate and cannot be seamlessly extended to bivariate case, not to mention design challenges with system's dimensionality, higher than bivariate. Presented investigation has proven that even with a limited underlying dataset, it is still feasible to appropriately predict system's failure/damage structural risks. Multidimensional dynamic CO2 transport/storage systems have to be designed safely, even based on a limited amount of underlying system's data. The proposed novel multivariate Gaidai structural hazard evaluation method had been validated versus the bivariate four-parameter Weibull-type method. Generic multimodal risk evaluation approach, benchmarked in the current study, may be applied to a range of complex dynamic structural systems, especially at structural design stages. The synopsis of this case study aims at contributing to future CO2 emissions reduction, which is believed to be beneficial for global environmental protection.