ORCID

Abstract

Oscillating Water Column (OWC) devices represent one of the most operationally robust wave energy conversion technologies, yet their commercial deployment remains hindered by narrow resonant bandwidth, geometry-dependent efficiency, and uncertainties in scaling thermodynamic effects between model and prototype. This thesis addresses these limitations through a coordinated investigation combining physical experimentation, empirical and machine-learning prediction, and high-fidelity computational fluid dynamics.A systematic experimental campaign was conducted on five single- and dual-chamber OWC geometries (Conventional, U-OWC, L-OWC, DCOWC-1, and DCOWC-2) under regular waves, with controlled compressible and incompressible chamber configurations. The thermodynamic scaling discrepancy is quantified for the first time across hybrid geometries, with peak efficiency errors reaching 43 % when air compressibility is neglected under high-damping conditions. Once compressibility is correctly scaled, the dual-chamber hybrids reveal the superior potential of geometric hybridisation, with the DCOWC-2 achieving a peak total Capture Width Ratio (CWR) of 0.85.The experimental evaluation is then extended to irregular wave conditions, encompassing both unimodal JONSWAP and bimodal sea states. These tests characterise the response of the investigated OWC configurations under controlled stochastic wave forcing and enable comparison of their performance across different spectral compositions. Among the tested geometry–PTO combinations, DCOWC-2 exhibited the highest and broadest total CWR response under the investigated irregular-wave conditions. The DCOWC-2 reached a peak total CWR of approximately 0.78 under unimodal seas and maintained a total CWR of approximately 0.60–0.72 across the tested bimodal conditions. The experiments therefore provide a controlled dataset for comparing OWC performance under unimodal and bimodal wave spectra.Building on these experimental findings, a unified predictive framework for U-OWC performance is developed by synthesising 518 data points from the literature and the present campaigns. Four modelling architectures, namely multiple linear regression, polynomial non-linear regression, decision trees, and random forests, are compared. The random forest ensemble achieves R² = 0.93 on the unified dataset, providing a rapid preliminary-design tool that bypasses the cost of new experimental campaigns.Finally, the air-spring effect in U-OWC devices is examined through a comparative incompressible-versus-compressible CFD study. Two configurations are evaluated: a geometrically restricted blanking setup and a thermodynamically scaled expansion-chamber setup. In the blanking configuration, the incompressible solver underestimates the resonant CWR by nearly 12%, while the compressible formulation substantially reduces this error. In the expansion-chamber configuration, interFoam instead overestimates CWR by up to approximately 30%, whereas compressibleInterFoam reproduces the resonant physical CWR with an error of less than 4%. These findings demonstrate that solver error is strongly configuration dependent and that explicit treatment of pneumatic compressibility becomes increasingly important as air-volume compliance and air-spring effects become significant. Together, these contributions provide a coherent body of experimental, numerical, and data-driven evidence for the design, modelling, and certification of broadband OWC breakwater systems suitable for next-generation marine renewable energy deployment.

Awarding Institution(s)

University of Plymouth

Supervisor

David Simmonds, Keri Collins, Hasan Gokhan Guler

Keywords

Oscillating Water Column, Wave Energy Conversion, Marine Renewable Energy, Coastal Engineering, Wave–Structure Interaction, Computational Fluid Dynamics, Experimental Hydrodynamics

Document Type

Thesis

Publication Date

2026

Embargo Period

2026-10-12

Deposit Date

October 2026

Creative Commons License

Creative Commons Attribution-NonCommercial 4.0 International License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

Download

Share

COinS