Understanding and including the dynamics of extreme natural hazard event uncertainty within the overall offshore wind farm project risk assessment using a causality-based graphical modelling approach

Abstract

Offshore wind structures are subject to the combined action of wind and wave loads. A change of these loads may significantly affect the integrity of the structural elements. Increased instabilities in the Earth’s climate system could increase the frequency of extreme events (e.g. rogue waves) well beyond the frequency values currently recommended within structural design standards. Inherent to extreme event modelling is the need to use expert (subjective) judgement and sparse data sets. In this context, a Bayesian Belief Network (BBN) can be applied to describe the effect of these changes on the frequency of rogue waves within wind farms located in shallow water depths of 20–60 metres. This graphical modelling approach provides the structure to effectively communicate, among others, parameter uncertainty, causality across multiple risk factors, quantitative definition of assessment subjectivity or potential impact of a change in rogue wave frequency relative to that described in current design standards. them in the customary Risk Assessment process of a company that operates physical assets in an offshore environment is entirely justified, despite its complexity and the high number of uncertainties involved. In the present work a causality-based probabilistic graphical modelling methodology is proposed to assess the risk associated with rogue waves in offshore wind farm projects at the final design stage. The methodology includes the impact of future climate change and provides the structure in which to effectively communicate: a) parameter uncertainty; b) correlation across multiple risk factors (i.e. “Systems of Systems” (SoS) complexity mapping/analyses); c) definition of assessment subjectivity; d) and potential impacts of low probability catastrophic events (i.e. extreme events). The methodology provides a holistic framework that can be integrated into existing decision-making processes currently defined within a large capital project execution process. In brief, the method studies the probability of a rogue wave impacting an offshore structure situated in a predefined location of the Northern North Sea, between 20 and 60 m depth, and includes 3 main stages: risk understanding, qualitative bow-tie creation; and transformation to a Belief Bayesian Network.