Tidal stream energy can contribute significantly to the future energy mix in the UK if harnessed using large-scale tidal arrays. In order to facilitate grid integration, the time variation of the power output from tidal farms must be assessed. A conceptual channel scale study is developed for a tidal farm consisting of 100 turbines deployed evenly in 4 rows with lateral and stream-wise spacing of 2 and 8 turbine diameters, respectively. The interaction between tidal farm and resource is simulated and both the mean velocity through the channel and the annual energy yield of the farm are reduced by 7.6% and 12.8%, respectively, due to the aggregate thrust from the turbine farm. For an array of identical turbines the capacity factor is 11% and the power output varies by up to 39% of the rated power. The occurrence of power variations greater than 10% and 25% of rated power is reduced by 13% and 37%, respectively, when a different operating point is assigned to each row of turbines.
A RANS-BE model is employed to further assess the influence of turbine operating point on array power production and support structure on net array drag. The RANS-BE model is evaluated against two single device rotor geometries and shows good agreement with an accuracy of more than 95% compared to experimental and numerical studies. The impact of an additional support structure immediately downstream of two rotors deployed side-by-side is subsequently assessed. The support structure is modelled as an actuator cylinder with an experimentally obtained drag coefficient of 1.09. The addition of a second rotor as well as the support structure results in an increase of the wake velocity deficit of 16%. This demonstrates that support structures can have a significant impact on the downstream flow features. The influence of this on array power production is subsequently assessed as part of a wider study on modelling power variation of tidal farms.