Modelling and prediction of blockage effects on the operation of tidal turbines
Koh, Martin Wei Xiang
Date of Issue2017-07-25
Interdisciplinary Graduate School (IGS)
Energy Research Institute @ NTU (ERI@N)
A series of works have been performed to better understand the effects of channel blockage on the performance and wake recovery of tidal turbines. The prediction of the performance of tidal turbines operating under blocked conditions has also been studied. Computational fluid dynamics (CFD) studies of a single turbine operating under various flow conditions and channel sizes revealed that the effect of blockage is to cause the turbine to experience a higher velocity than the free-stream velocity. The boundary layer is also found to greatly affect the prediction of performance of a tidal turbine, and neglecting its presence would lead to the performance of the tidal turbine being overestimated. Channel depth is also found to affect the performance of the tidal turbine. Increasing the blockage ratio of the channel is found to increase the recovery of the far wake, while the near wake was minimally affected. Predictions of performance by analytical models were found to be more sensitive to changes in blockage ratios when compared to that predicted from CFD. Nevertheless, there is generally good agreement between the predictions made by analytical models and CFD for typical operating ranges of thrust coefficients. Prediction of the wake recovery was also compared. When modified to work under blockage conditions, predictions of the wake recovery by analytical models had good agreement with the wake predicted by CFD. The model used by Bahaj et al. in 2007 to predict the effect of blockage on the performance on a turbine, together with the Jensen wake model, is recommended for predicting the effects of blockage on the performance and wake recovery of a single turbine. Coupling the blockage prediction model by Bahaj et al. with a BEM model allowed the prediction of the performance of a 3 bladed rotor under blockage conditions with a good degree of accuracy. In addition, Reynolds number effects are found to be important in accurately predicting the performance of the turbine under blocked conditions due to the increase in experienced velocity. Including the Reynolds number effect results in the optimum rotation speed and power coefficient increasing as the blockage ratio increases as compared to when it is not included. The performance and wake recovery of a small array of fixed size under various blockage scenarios was also studied. Performance of each turbine within the array is found to be between that of a single turbine operating with the same global blockage and local blockage as the array. The equivalent single turbine blockage is defined as the blockage ratio required for a single turbine to produce the same performance as the turbine within the array. The equivalent single turbine blockage is used to characterise the performance of the turbine in the array, and an expression was developed to describe it. Wake recovery of the array was studied as well. Increasing global blockage ratios was found to increase the wake recovery of the array, while locating turbines too close to each other lead to their wakes merging and taking longer to recover. When there is no significant wake interaction between the wakes of the individual turbines in the array, the wakes were found to be similar to that of a single turbine operating with the same global blockage ratio as the turbine. The minimum inter-turbine spacing for this is found to be two turbine diameters.