Investigation of updating methods for probability-informed inspection planning for offshore structures

Abstract

The paper presents a novel process for probability-informed inspection planning based on the following principles. 1) A description of the strength of knowledge about a problem shall be included and be accounted for in the decision making, 2) Probabilistic methods should be used to predict quantities expressing the physical reality of nature (observable quantities) and their uncertainty, and 3) Any update of the probabilistic model should be limited to the actually observed quantities. The process for updating the probability of failure after inspection is programmed in accordance with these principles based on Monte-Carlo simulations and Bayesian parameter updating. The application of these principles and the proposed process is illustrated by an example calculation resulting in an example of inspection intervals for a jacket structure.