Investor confidence in the wave energy industry has been eroded by the collapse of forerunners who went “too big too quickly” and relied on models that were developed primarily for other industries such as Oil & Gas. The underlying assumptions of linear wave theory may be valid for the analysis of large static structures but it proves inaccurate when investigating the behaviour of small dynamic systems, particularly those with geometrical asymmetry and deployments in shallow water. CFD is able to deal with these complex fluid-structure interactions by solving the governing equations of the dynamics of the fluid. However, it is difficult for a developer to justify the additional computational expense and expertise required to accurately implement CFD unless its power and value are fully exploited. In this thesis OpenFOAM is used to develop a numerical wave tank, along with a fully integrated non-linear PTO model, to allow investigation of the whole CCell system, including the coupled behaviour of the hydrodynamics and PTO system. In a comparative exercise evaluating hydrodynamic coefficients, the CNWT demonstrated the information that would be overlooked if a linear model, such as NEMOH, was instead adopted. An increased understanding of the wave-structure interaction is provided by the CNWT, allowing interrogation of CCell’s defining parameters, which lead to the pilot deployment design increasing its CWR in predominant wave conditions by 69.8% in comparison to an equivalent flat paddle. Furthermore, conveyance of the pressure distribution from extreme loading scenarios facilitates the structural design of the composite WEC through FEA software analysis. The coupled PTO model also provides representative operational flow rates through the hydraulic system, which allows better characterisation of the PTO system in dry testing. Combination and adaptation of these different numerical tools can provide an efficient modelling strategy for the development of a WEC, yielding a more accurate understanding of the system dynamics and overall performance than commonly used linear codes. Increased confidence in these model outputs accelerates the design process for developers and ensures that the concept is more thoroughly tested and de-risked prior to expensive prototype demonstrations. Adoption and extension of the numerical strategy as detailed within this work could lead to renewed credence and, more importantly, investment to the industry as a whole.