The primary function of seawalls and embankments is to protect against damage and injury caused by flooding. Coastal flooding is caused by combinations of high tides, waves, wind set-up and storm surges driven by low-pressure systems. However with global warming causing sea levels to rise and with increased storminess causing more extreme waves and storm surges, the likelihood of overtopping of seawalls with zero or negative freeboard may well be expected to increase. Researchers using physical and numerical models to develop design formulae have widely investigated wave overtopping of seawalls with positive freeboard. However the design of seawalls with zero or negative freeboard has attracted much less attention, and some variation exists between overtopping discharge calculated with current design formulae. The focus of this thesis is the extreme situation when overtopping caused by storm waves is combined with surge levels above the embankment crest. The local highly accelerative flow over the embankment crest caused by the high surge level will significantly alter the flow at the crest. This is likely to have a highly non-linear effect upon the overtopping waves. In this thesis, the flow is investigated with a 2DV numerical model based on the Reynolds averaged Navier-Stokes (RANS) equations developed by Lin and Liu (1998a). The model describes the flow characteristics of a breaking wave such as the velocities within the wave as well as the turbulence at the seabed boundary layer. As an example of the model’s ability to describe complex hydrodynamic flows, this study investigates its ability to represent the second order mass transport under progressive and standing waves. The model results are compared with available theory and experimental results. This shows that mass transport is successfully predicted, although there is some variation in the magnitude compared to the experimental and theoretical results. To consider the model’s ability to simulate storm surge wave overtopping of embankments, the RANS model has been used to simulate an experimental study conducted by Hughes and Nadal (2009). To examine the success of the model at reproducing the wave generation, transformation and overtopping processes the model results have been compared with the experimental laboratory data. This makes possible a wave-by-wave comparison of overtopping parameters such as discharge, depth and velocity for a storm surge event. Additionally the overtopping discharge predicted by the model is compared with design formulae and the differences in the overtopping discharge calculated with current design formulae are investigated and explained. Finally, the RANS model is used to determine the effect of embankment crest width on the magnitude of the overtopping discharge. Results from RANS model tests are used to provide design guidance in the form of an equation that allows the effect of crest width to be included when evaluating combined discharge at embankments.

Recent hurricanes in the Gulf of Mexico (e.g., Ivan, Dennis, Katrina, Rita and Ike) were observed to develop wave conditions that were near or exceeded the predicted 100-year conditions. As a result, many offshore facilities, as well as coastal infrastructure, which were designed to withstand the 100-year condition, were damaged. New estimates of extreme conditions, which incorporate recently observed maxima, are needed to provide better guidelines for design of coastal and offshore structures. Berek et al. (2007) have used modeled data to develop new criteria, but these estimates can be very sensitive to the data and to the statistical methods used in the development. Berek's estimates also do not cover the entire Gulf of Mexico. We have developed updated estimates of the 100-year extreme wave conditions for the entire Gulf of Mexico using a more comprehensive approach. First, the applicability of standard parametric wind models was examined and appropriate adjustments to the Rankine vortex model were developed to reduce the wind field errors during hurricane conditions. The adjusted winds reduced the error by up to 25 percent compared to the original Rankine vortex model. To obtain reliable wave data, merged wind fields were generated using the NCEP/NCAR Reanalysis 1 project modeled wind data for background wind and the parametric wind model for hurricane conditions. Next, the SWAN wave model was used for the 51-year period from 1958 to 2008 along with multiple statistical methods (Gumbel, Weibull and GEV-Generalized Extreme Value distribution). The effect of the recent hurricane season (2004-2008) shows that maximum 100-year wave height values and their distribution changes. A resampling technique (bootstrap) is used to evaluate and select the optimum statistical method to estimate more appropriate extreme wave conditions.

This thesis is concerned with the use of the open source computational fluid dynamics (CFD) software package, OpenFOAM® for predicting and analysing the behaviour of a near-shore oscillating wave surge converter (OWSC), when subject to various types of ocean wave conditions in a numerical wave tank (NWT). OpenFOAM® which utilises a Finite Volume Method (FVM) is used to solve the incompressible, Reynolds Averaged Navier-Stokes (RANS) equations for a two-phase fluid, based on a Volume of Fluid (VOF) phase-fraction approach to capture the interface between the air and water phases. Preliminary studies on classic wave-structure interaction benchmark cases, involving a fixed and a vertically oscillating semi-immersed horizontal cylinder are carried out. The gradual transition of the linear to non-linear behaviour of the horizontal and vertical forces induced on a fixed cylinder when subject to various regular waves, and the amplitude ratios of the surface waves elevations generated by the prescribed oscillatory motion of the cylinder, are shown to provide good overall agreement within the limitations of the relevant theory and the experimental data in the literature. The OWSC is modelled with the inclusion of a Power Take-Off (PTO) system, using a linear damping restraint, and simulated in two-dimensional (2D) and three-dimensional (3D) setups. The 2D and 3D numerical results, such as the surface wave elevations, flap angular velocity, PTO torque and flap angular displacement, compare well with one another and with the experimental data for operational regular head-on and oblique wave conditions. Small discrepancies between numerical results and experimental data are likely to be caused by non-linear behaviour of the PTO system. Pressure distributions on the flap surfaces and forces induced on the flap and hinge of the OWSC for various wave conditions are also presented. The effects between 2D and 3D wave-structure interactions become more significant when subject to large waves that break during impact. Comparison between the full scale and 1:24 scale numerical results of the OWSC shows no significant evidence of viscous and scaling effects. The validated 2D OWSC model is also subject to embedded focused waves, to predict the worse possible scenario of wave loading in extreme wave conditions. The delay of the focus event breaking is shown to affect the slamming behaviour for the larger focus event wave heights. Incorporation of a focused wave at different phase positions within a background of regular waves reveals that the focus event wave height has little effect on the peak tangential force on the flap during the slamming event, when a PTO cut-off mechanism is implemented to prevent excessive torque surges. In contrast, the peak radial force on the flap and the maximum resultant force on the hinge appear to respond more sensitively to the focus event wave height. It has been demonstrated that OpenFOAM® is able to provide a comprehensive understanding of the complex hydrodynamic analysis and prediction of highly non-linear wave-structure interactions for an OWSC, which give useful guidance and confidence to WEC developers on the design considerations relevant to the OWSC systems.