Using Multibeam Sonar to Characterise the Wake of a Cross-Flow Tidal Turbine

Abstract

Tidal energy is a more predictable and smaller footprint alternative to wind energy, while crossflow turbines enable deployment in shallower waters. As with wind energy, it is important to optimise the array performance of such turbines by understanding their wake development and recovery in real conditions. While particle image velocimetry (PIV) has been used to characterise the wake of model turbines in flumes, the high turbidity in field studies requires the use of SONAR based methods instead, like acoustic doppler velocimetry (ADV). But ADV can only produce point data, which is limiting in highly temporally varying flows. It can also be difficult to position such devices in a way that will not alter the flow, requiring a supporting structure in the region of measurement.

In this paper, a new multibeam SONAR method was tested for providing a fuller picture of wake characteristics in real, highly turbid environments without interference in the local flow. Two devices were trialled - the Kongsberg EM2040P and the Oculus M3000d - and the higher frequency Oculus (1.2 MHz) was judged to be more suitable for this application. Different seeding techniques using bubbles and particulate where also trialled, but seeding using bubbles proved more convenient in the field than particulate.

A field campaign was conducted to gather time-series of 2D cross-sections of the wake behind a full-scale cross-flow tidal turbine. Preliminary analysis clearly demonstrated that the method was successfully able to capture features of the flow. Postprocessing of the data is ongoing and aims to identify discrete particles in the images which can be tracked to reveal the velocity field of the wake.

This method requires further development, but initial results suggest it will be valuable in future studies of tidal turbine wakes, as well as other hydrodynamic applications.