A Risk Assessment Framework for Offshore Wind Turbines
David Wilkie, University College London
Offshore wind turbines (OWTs) must be sufficiently resilient to withstand both storms and operational deterioration over a 20-25 year life in an aggressive marine environment. Extreme storms, such as Typhoon Rammasun in 2014 (a Category 5 super typhoon with winds of up to 259 km/h) caused failure of the structural components in onshore wind turbines. However, similar empirical observations do not yet exist for OWTs. In addition, OWTs are rapidly increasing in size: the 9MW OWT introduced last year by Vestas has a rotor diameter of 154 meters, more than twice the diameter of the earliest OWTs. This means that observed failure rates from previous generations of OWTs would not be suitable for the current, much larger, generation. Current risk assessment procedures for OWTs often neglect structural failure and focus on the equipment only, which can be assessed using existing empirical databases. However, this is not enough: an OWT is a complex integrated system where failure of the structure may cause loss of all equipment. This project developed a catastrophe risk modelling framework to assess the risks associated with offshore wind infrastructure exposed to extreme wind and wave conditions, to ensure safe designs and to price insurance policies. A site-specific assessment of structural fragility is developed, to quantify the probability of structural failure at different environmental conditions using aero-elastic analysis. Separately, a hazard model defines the probability of different environmental conditions occurring. Combining the hazard and fragility enables the estimation of structural failure rates which can then combined with empirical mechanical and electrical component failure rates to assess financial losses of an entire OWT system. Additionally, degradation in the form of small cracks that grow from weak points in the structure over many years of energy production and eventually threaten the integrity of the machine (fatigue) is modelled. However, numerical assessment of fatigue damage over the life of a structure is computationally expensive, due to the need for aero-elastic simulation of a large number of environmental conditions. This makes structural fragility for FLS a challenging task as it also requires numerical sampling of random variables to model uncertainty in the estimation of fatigue damage. This project proposed using Gaussian process regression to build surrogate models for fatigue damage caused by different environmental conditions. Ultimately, a structural reliability calculation using the surrogate model highlights the large scatter in fatigue life prediction due to parameter uncertainty, and enables the calculation of loss functions. The results from a case-study application in Dutch waters indicate that failure of the structure plays a major role on the overall risk profile of an offshore wind farm, and that for an OWT located in European waters fatigue dominates the structural financial losses.