Wake and cluster research: past, present and future

The feasibility of constructing large arrays of wind turbines (100¿¿1000 machines) has now been firmly established. The energy losses are predictable, and can be kept below 20¿¿25% with suitable choice of spacing. The energy yields of clusters have been calculated so as to quantify the UK offshore windpower resource, and estimates have also been made as an input to assessments of the problems and costs of both onshore and offshore siting. Plans for construction of arrays have now been announced by a number of electricity supply authorities, including the CEGB. Provision of performance prediction data for these arrays, particularly the smaller ones, will necessitate development of existing techniques to take into account the operating characteristics of particular machines. The procedures are illustrated using the specific example of the Nibe cluster, where wake measurements are currently being made, and the annual energy loss for this cluster is estimated at 3%. It is concluded that spacings can be reduced with fewer machines: from the 10¿¿12 rotor diameters applicable in large arrays down to five diameters, as at Nibe, with, say, two-four machines. The requirements for further analytical and experimental studies are discussed, with particular reference to the current UK programme, but also, where appropriate, to studies being carried out elsewhere. Data on the wakes and interactions of large machines are needed, especially for comparison with estimates derived from the predictive techniques. An assessment of the limited data currently available confirms, qualitatively, the value of corroborative studies at model scale, but points to the need for a more detailed understanding of the mechanisms of wake generation and decay.