Gaidai Risk Evaluation Method for Lifetime Assessment for Offshore Floating Wind Turbine Gearbox

Abstract

Renewable clean energy in some cases may be viewed as an alternative to limited fossil resources. Offshore floating wind turbines (FWTs) are among the most attractive green alternatives. However, FWTs, in particular their essential components, may sustain structural damages from cyclic loads brought on by torque, bending, longitudinal loadings, as well as twisting moments. Multibody simulation tool SIMPACK was utilized to assess structural bending moments and internal forces occurring within the FWT drivetrain during its field operation. The novel risk and damage evaluation method advocated in the current study is intended to serve contemporary FWT design, enabling accurate assessments of structural lifespan distribution, given in situ environmental/field conditions. The approach described in the current study may be utilized to analyze complex multidimensional sustainable energy systems, subjected to excessive stressors during their intended service life. Contemporary risk evaluation approaches, dealing with complex energy systems are not always well-suited for handling dynamic system's high dimensionality, aggravated by nonlinear cross-correlations between structural components, subjected to dynamic nonlinear nonstationary loadings. The current study advocates a novel general-purpose lifetime assessment methodology, having a wide area of potential engineering and design applications, not limited to offshore wind/wave renewable energy systems. Key advantages of the advocated methodology lie within its robust ability to assess damage risks of complex energy and environmental systems, with a virtually unlimited number of system components (dimensions), along with the further potential to incorporate nonlinear cross-correlations between system components in real time. Note that to the author's knowledge, there are no comparable risk evaluation methods that can deal with the system's high dimensionality, utilizing raw/unfiltered simulated/measured datasets, beyond one or two-dimensional dynamic systems—except for computationally expensive direct Monte Carlo (MC) simulations.