In the present paper, two types of wave energy converters (WECs) in uni-/multi-directional waves are investigated using a potential-flow model. The first example is a flap-type oscillating wave surge converter (OWSC) which is similar to the configurations of Oyster wave power device. The second case is an attenuator-type WEC which is based on the Pelamis P2 machine; the modelled WEC is simplified as 5 interconnected rigid modules. Both hydrodynamic interactions and mechanical connections have been considered in the present analyses. The emphases have been put on the effects of directional spreading on the performance of the WECs. Significant reductions of power output have been found in multi-directional seas.
There have been many designs or concepts to harness wave power from the ocean. Wave energy converters (WECs) can be divided into different groups according to the method used to capture the wave power (http://www.emec.org.uk/marine-energy/wave-devices/), i.e. attenuator, point absorber, oscillating wave surge converters (OWSC), oscillating water column (OWC), overtopping/terminator device, submerged pressure differential, bulge wave, rotating mass, etc. Most wave energy devices convert kinetic energy into mechanical motions to generate electricity. They usually include moving components and complicated mechanical conversion systems (e.g. power take-off systems). So both hydrodynamic and dynamic interactions have to be considered in numerical investigations.
Numerical methods for hydrodynamic modelling fall into potential- or viscous-flow frames (Li and Yu, 2012). Linear or nonlinear wave theories can be used for potential-flow analyses. Linear wave theory is usually used for operational sea states and nonlinear wave theory is adopted for the analyses of strongly non-linear waves and extreme events (Coe and Neary, 2014). When viscous effects cannot be neglected, empirical viscous coefficients have been introduced to provide reasonable predictions. However, these coefficients are geometry dependent and limited to model scale. The difficulty of this approach has been highlighted by Pauw et al. (2007) and Sun et al. (2015). A good alternative is to use a computational fluid dynamic (CFD) model such as that presented by Wei et al. (2013).
