Enhanced array design for tidal power generation

Tidal stream energy is a predictable source of renewable energy. Tidal stream turbines have been proposed as a way to extract useful energy from the tide. Many arrays of such devices will need to be installed to extract significant amounts of energy. The presence of an array of turbines within a tidal flow will impact the flowfield, as complex fluid interactions occur across multiple scales. This thesis is concerned with the behaviour of tidal turbines arrayed across channels. Experimental and analytical work is carried out to investigate array behaviour and to create new modelling tools to replicate this behaviour. Linear Momentum Actuator Disc Theory (LMADT) is employed to develop a new analytical model for a long row array of tidal turbines split into multiple smaller, co- linear row arrays. An argument of separation of scales is used to facilitate this model. It is found that increases in power extraction beyond that of a single continuous row array are possible. Experimental work is carried out on a row array of eight porous discs, simulating a short row array of tidal turbines. Disc porosity and spacing are varied to investigate thrust on the array, flow behaviour behind the array and an 'inferred' power removed from the flow. The results are compared to previously developed theoretical models. Good agreement is found with the trends of the analytical model, for example that there is a peak power coefficient which can be reached through appropriate selection of spacing and disc resistance. Differences from theory are found in the total thrust and power measurements, as well as in some aspects of the flow behaviour in the array wake. Reductions in thrust and power towards the ends of the array are also identified as 'end effects' which are not included in the analytical model. Based on these results a new semi-empirical model is proposed, using LMADT with experimental data closure. This model allows variation of the disc resistance across a row array. Values from the experimental work are used as inputs to the model, and the results compared to experimental measurements of flowspeed, thrust and power. Although agreement with experimental results is found in some areas, there are still some discrepancies between the analytical model and the experimental results. This indicates that there are additional factors that contribute to end effects on a short row array.