FPSO hull structural integrity evaluation via Bayesian updating of inspection data

Preserving the integrity of any system depends on predicting, assessing, and preventing risks while any existing failure mechanism is mitigated. This depends on forecasts, technological innovations and finding appropriate solutions to prevent or deal with emergencies. The organizational structures that understand and manage their risks have increased process stability and operational availability. To achieve this objective an accurate estimate of decision analysis parameters based on risk is needed, but this requires significant amount of data. When the observed data are limited, statistical estimates can be used to supplement or even replace information based on the Bayesian approach. This paper applies these concepts to the study of the structural integrity of the hulls of floating production storage and offloading (FPSO) units. Two case studies were conducted with a focus on corrosion damage due to its high incidence among the observed damages. Approximately half of the damage cases observed in practice are due to corrosion degradation. With the results from Bayesian inference, the necessary subsidies are developed for the next field inspection, establishing the basis for determining the regions where survey is needed the most, and the corresponding inspection intervals.

► More inspection campaigns more precise estimates.
► Posterior distribution similar to real distribution.
► Not updated repair lead more severe results than actual.
► Bayesian inference provides coherent estimate.