• BEM model
• DMST
• farm optimisation
• HATT
• MIQP
• SHYFEM
• VATT

The focus of this thesis is the development of a turbine model within the Shallow Water Code SHYFEM, to perform energy and environmental impact studies. SHYFEM (System of HydrodYnamic Finite Element Modules) is an open access software developed at the National Research Council-Institute of Marine Sciences (CNR-ISMAR) of Venice. It was born for oceanographic purposes, and is therefore able to reproduce waves, tides, bathymetry, coastal morphology, and other aspects as they occur in realistic situations. In the first part of the PhD research project, we developed the ”in-house” turbine model to be embedded in the SHYFEM code and used for energetic analyses. The turbine model is 3D and has been developed for both Horizontal and Vertical Axis Tidal Turbine (HATT and VATT). The model has a momentum sink approach: we have no blades in the grid. The presence of the turbine is reproduced by introducing momentum sink terms in the x and y momentum equations (since the code is hydrostatic, i.e. no z momentum equation is present, while vertical velocity is computed from the continuity equation). The value of such sources is based on the forces that would act on the blades if they were present. The forces are calculated using the Blade Element theory. We calculate the forces that would act on a blade element and then apply equal and opposite forces to the flow. In this way, it is possible to reproduce the presence of the turbine with very little computational effort. Such a model has been extensively validated against experimental data. The second part of the research aims to use the developed code for farm layout optimisation. We focus on fluid dynamics to maximise power generation. This is essential to reduce the Levelized Cost of Energy and thus enable the commercial development of tidal energy converters. To this end, we use a simple Mixed Integer Quadratic Programming (MIQP) model to optimise the farm layout using a discrete approach. The MIQP model is equipped with an analytical model, taken from the literature, to describe the wake development. The purpose of this second part is to show the potential of such an approach as well as its limitations. In particular, we demonstrate that further improvements are needed to take advantage of favourable flow conditions that may arise within a cluster of turbines, but are not accounted in the analytical wake model. To overcome these limitations, Computational Fluid Dynamics (CFD) simulations should be used as an optimisation tool. Thus, we needed to remove the bottleneck of limited CFD simulations due to manual computational grid generation and setup. So, we automated the process. With appropriate computing resources, this could be the starting point for a Machine Learning (ML) approach to tidal farm layout optimisation. This will also allow the user to have a wide available casuistry to be analysed for a deep understanding of basic fluid dynamic mechanisms occurring between adjacent devices. Moreover, the characteristics of the SHYFEM code allow for site-specific analyses: once the tidal farm has been adapted to the site, i.e. the layout has been optimised, SHYFEM can be used to predict the energy yield and assess the impact on the environment.