Offshore renewable energy systems : Quantification of extreme loads using computational methods

• This Ph.D. thesis investigates the dynamic response of offshore energy systems in extreme waves. The use of offshore energy technologies, such as wave energy systems and offshore wind turbines, is crucial for transitioning to clean energy and mitigating the effects of climate change. However, to design reliable systems, it is important to understand their behavior in harsh environmental conditions.The first part of the thesis focuses on classical Computational Fluid Dynamics (CFD) simulations for modeling the response of structures in extreme waves. Breaking waves are numerically reproduced and the corresponding slamming loads are estimated, as well as the maximal forces on critical components such as the mooring system. The thesis addresses the challenge of computational mesh deformation, which can lead to numerical instability and failure in simulating extreme structural responses. Dynamic mesh techniques are implemented to overcome the limitations of classical techniques. Additionally, the thesis explores alternative approaches to representing a sea state, such as equivalent regular waves and focused waves, to reduce the computational cost of full sea state simulations. A mid-fidelity numerical model is also employed, with its accuracy verified against a high-fidelity solution.The second part of the thesis advances the use of probabilistic machine learning to develop a surrogate model for the mapping between extreme waves and the corresponding forces on the structure. A Bayesian active learning method is employed to train the model with high prediction accuracy, especially in extreme events. The surrogate model is many orders of magnitude faster than classical modeling methods and enables efficient statistical quantification of the quantities of interest, such as loads in critical system components.Overall, this thesis provides a comprehensive examination of advanced computational methods for estimating the dynamic response of offshore energy systems in extreme waves and enables reliable and cost-effective design through the use of fast and accurate surrogate models.

Ämnesord

TEKNIK OCH TEKNOLOGIER  -- Maskinteknik -- Strömningsmekanik och akustik (hsv//swe)

ENGINEERING AND TECHNOLOGY  -- Mechanical Engineering -- Fluid Mechanics and Acoustics (hsv//eng)

Nyckelord

offshore systems

wave energy

CFD

extreme events

machine learning

Bayesian experimental design

active learning

surrogate modeling

Publikations- och innehållstyp

vet (ämneskategori)

dok (ämneskategori)