Data-Driven Design and Operation of Offshore Wind Structures

Abstract

In this paper a novel Data-Driven Design and Operation approach is introduced for an offshore wind energy application to account for site-specific conditions. Data from operational state of a turbine, e.g. accelerations, strain gauges, and SCADA, are intended to be used in the True Digital Twin concept to improve initial design assumptions. A calibrated numerical model combined with measured loading and response models allows to update the structural reliability and thus quantify the potential inherent conservatism. Given this additional knowledge the lifetime extension of existing structures can be assessed while for future structures their design procedures can be optimized according to a chosen cost function. The conceptual framework for including new information from measurements based on the Bayesian updating is presented in the paper. It is demonstrated how the initially assumed generic uncertainties regarding numerical models and environmental conditions can be consistently updated based on the site-specific data and hence lead to a more reliable design and operation of a structure. New information derived from data is accounted for based on the pre-posterior decision procedures supplemented with robust decision rules.