Dynamic Modelling of Wind Loading on Floating PV sytems
More Info
expand_more
Abstract
As the global community seeks to improve current energy technologies for a sustainable future, offshore PV systems show great potential to overcome the land use limitations of traditional PV systems. There is a complex interaction between different environmental forces at sea with the PV structure. Empirical studies on these environmental interactions can provide critical insights to optimize the design of offshore PV systems. This thesis has identified a gap in wind loading studies on offshore PV designs. The primary objective of this research is to develop a methodological approach for studying the wind-induced tilt variations in floating PV (FPV) systems, with the intent of informing future PV performance assessment frameworks. A numerical modelling approach using the CFD tool ANSYS Fluent is adopted as the base methodology. The first phase of the study employs single-phase wind tunnel models in 2D and 3D to study flow behaviour over the required geometry. The models are validated against literature and used as the basis to construct a multi-phase model to simulate wind flow over a floating body and how it moves in response to the water surface.
The model is used to perform sensitivity studies on wind speed, module orientation, module tilt, and floater height. The simulation results indicate that tilt variations induced by wind flow are minimal, under 5◦, for still water conditions. In addition, it also suggests an increase in wind velocity increases the non-linearity in the rotational response of the body, changes in the direction of rotation depending on the speed of the flow and tilt of the module, and a sharp increase in rotation with an increase in floater height. The highest rotational response of 4.07◦(anticlockwise) was observed for a design tilt of 24◦ and a wind speed of 40 m/s.
Based on the simulation results, a PV yield assessment is done, and it is found that the inclusion of tilt effects changes the yield by less than 0.5%. Given the high computational cost involved, a numerical approach for wind load evaluation is not justified in terms of yield variation. For future work, it is recommended to first conduct experimental studies to verify the effects of wind loading on different FPV system designs. This could be supplemented with numerical approaches in the future using high-performance computational resources.