CYBERLAB aims at studying offshore power grids that take the form of a lattice of marine structures, i.e. a vast set of floating structures interconnected by a shared mooring system in a scalable and cost-optimal way. Present simulation tools, experimental methods, and offshore standards are primarily developed for single floating bodies. Numerous additional challenges arise when it comes to the design and analysis of lattices, which have not been addressed thoroughly yet. CYBERLAB aims at developing radically new (cyber-physical) empirical methods to understand the dynamic behaviour of such lattices. The utlimate goal is to make the Norwegian offshore community able to design optimal and competitive lattices, and become a first-mover on the large-scale green energy market. CYBERLAB will be headed by SINTEF Ocean, and gather international academic experts from NTNU and Aarhus University, supported by an outstanding team of industrial partners (Equinor, Aker Offshore Wind, APL, Sevan SSP and Deep Sea Mooring). The methods developed in CYBERLAB will be at the heart of the upcoming Ocean Space Centre in Norway.
Future ocean energy installations are changing in character, from single moored floating structures to large lattices of floating structures, such as floating wind parks or interconnected solar islands. Lattices are here defined as two-dimensional arrangements of identical structures (cells), interconnected by soft connections such as mooring lines.
Lattices have a more complex dynamic than monolithic structures (many more eigenmodes), and will therefore require an accurate modelling of excitation and damping mechanisms. To design and verify lattices, computational models that account for the reciprocal interaction among elementary cells of the lattice and the hydrodynamic loading must be developed, and validated with experiments. However, the limited size of existing hydrodynamic testing facilities would require performing tests at ultra-small scales that entail scaling issues.
The present project proposes to use cyber-physical (CP) empirical methods to address this problem, in which the dynamic system under study is partitioned into physical and numerical substructures that interact with each other in real-time through a control system. In CP empirical methods, the behaviour of the physical substructures is partly unknown, while the numerical subsystems are described by validated computational models. In marine technology, such methods have been developed to enable empirical studies on floating wind turbines and led to the discovery of unexpected effects, such as wind-velocity dependent eigenmodes.
Lattices will be studied by iteratively calibrating the parameters of the hydrodynamic models of the simulated cells based on experiments involving a unique physical cell, and simulating the rest of the lattice. Bayesian optimal experimental design (OED) will provide a mathematically sound framework for validation and calibration of hydrodynamic loading models in this context.