Offshore floating wave energy converters (WECs) are a new and rapidly growing technology, driven by the need to reduce CO2 emissions and provide renewable energy sources. The design process of such a device involves the use of computational tools and several stages of physical tank testing of scale models. Physical tank tests are generally expensive and can only provide insight into the physics taking complex scaling effects into account. Full scale models cannot be used, as WECs are generally large in dimension and thereby too large for controlled tank tests. Different computational tools exist, which are either linear or non-linear. Linear models, such as boundary element models, are used to describe the system of WEC-mooring-PTO-waves by means of a linear, predominantly frequency-based, descriptions and are traditionally used in control design. The associated assumptions of inviscous fluid, irrotational flow, small waves and small body motion, however, are a major limitation of this modelling approach, since WECs are designed to operate over wide wave amplitude ranges, experience large motions with waves breaking at the body and generating viscous drag and vortex shedding.
With consideration of the full range of effects, the physics can be described using the Navier-Stokes equations. This branch of computational model falls into the category of computational fluid dynamics (CFD). It is very computationally expensive and not suitable for use in the design process of WECs with regards to maximising the efficiency (through design iterations and development of control algorithms) in real sea states, where long-time simulations are required.
This project aims to combine the strengths both types of modelling concepts. In particular, the mathematical description of the technical system including the main device, its auxiliaries such as the mooring and power-take-off (PTO), and also the waves, that excite the device, is a necessity for the design of controllers but also when simulating the interaction of WECs in an array. Instead of using linear coefficients in this model, we propose to generate non-linear parametric models using system identification techniques, based on WEC responses obtained from numerical tank tests performed using CFD. These will take into account wave run-up in front of the structure, time varying degrees of submergence, which results in changes of buoyancy and restoring forces. It will also model viscous effects, such as vortex shedding and drag, and turbulence.
The project is divided into three main sections. Within the first stage, an experimental setup will be developed, that can be used in a numerical wave tank in CFD. Such an experiment would need to provide the data from which radiation effects due to the moving body, and diffraction effects due to waves exciting the WEC, can be identified. Possible solutions may include decay tests, where the body is released from unstable positions and the response is measured. For the fixed body, i.e. to get diffraction results, waves of varying frequency may be generated and the surface elevations, depending on the incoming wave, can be processed.
The next part of the project involves system identification. The CFD tests can be standardised and routines need to be developed that identify the necessary coefficients from the CFD results. The Centre of Ocean Energy Research at NUI Maynooth has considerable experience, over more than 20 years, in the development and application of system identification methods for a wide range of industrial applications. A crucial step in this stage involves the determination of appropriate model parameterisations which capture the essential nonlinear dynamics of families of WEC types. The parametric description can then be used in the third stage of the project to design a true nonlinear control algorithm to optimise energy conversion. The benefit of such controllers is that they respect the true nonlinear dynamics of WECs and are likely to give realistic energy maximisation in any irregular, non-linear sea.
Therefore, the project will provide an important step in extending the range of tractable computational techniques to the design and control of WECs, shortening the leap to expensive tank testing and providing realistic mathematical models upon which to base energy-maximising control designs.
The final outcome of this project will be a modular software suite, which can be licensed to end users. Furthermore, a company will be set up, that can provide the proposed technology as a service. The company can charge for the CFD model, system identification routine, development of the motion model and controller design separately. Also, as the CFD model is already available, a final long-time full-scale simulation with the controller in place can be offered. This would be similar to the full scale device being deployed in controlled conditions.