Abstract
Monitoring marine ecosystems and maritime activity in the North Atlantic requires autonomous sensing platforms that can operate for long periods with limited maintenance. Conventional power sources like batteries have finite lifetimes, while solar panels and externally exposed generators are limited by winter darkness, salt buildup, corrosion, biofouling, and mechanical damage. This thesis investigates a sealed internal wave-energy harvester that converts buoy motion into electrical power while keeping the main energyconversion mechanism protected inside the buoy. The selected architecture is a linear permanent-magnet electromagnetic harvester in which a coil assembly translates along a segmented magnetic rod.
The concept is developed through electromagnetic modelling, prototype construction, experimental characterisation, coupled electromechanical validation, and forward scaling analysis. A bench-scale prototype was built with a three-coil carriage and onboard instrumentation for measuring electrical power and carriage motion. A custom rope-driven wave simulator was then developed to provide repeatable vertical excitation for controlled testing. Prototype testing and coupled electromechanical modelling were used to identify mechanical losses, electromagnetic damping, resonance behaviour, and generated power, providing a calibrated basis for evaluating the architecture under larger-scale design assumptions.
Scaled-design simulations based on the validated model then examined whether the architecture could approach the long-term target of 10 W average electrical power for a compact, man-portable monitoring buoy. The simulations used representative sea states from a quality-filtered ocean wave record, with buoy excitation approximated using a simplified wave-follower assumption. The model-based results indicate that reaching the target is technically plausible but demanding, requiring increased moving mass, improved electromagnetic coupling, careful stroke management, and load-tuning strategies. Within a strict 1.8 m rod-length envelope, the best scaled candidate predicted an output of about 8.8 W under the screening assumptions, while larger packages reached the 10 W class. The results support further development of the selected architecture, with the main remaining challenges being a practical stroke-management mechanism, mechanical integration, and further buoy-coupled hydrodynamic modelling.