Session: 09-07-01 Tidal & Current Energy II

Submission Number: 176346

Use of Stochastic Estimation for the Wave-Turbulence Separation in the Water Column 

The Alderney Race (in France) constitues one of the most  tidal-stream energy potential in the French coastal sea. In such an area, there is a complex interaction between wavy environment and large scale flow structures that develop in the water column and originate from irregular bathymetry sea floor. The characterisation of the tidal turbine power fluctuations resulted from this interaction is a great interest to improve the performance of a tidal turbine farm. Before trying to understand the main source of these power fluctuations, one of the great challenging task consists in elucidating the effects of the waves on the large scale flow structures developing in the water column, especially the associated energy transfer.

Such an interaction has been previously investigated by trying to extract the wave component embedded in the raw velocity fields by using for instance Wavelet decomposition, Empirical Mode Decomposition or Dynamical Mode Decomposition. But each of these methods is dependent on the underlying principles of decomposition and requires as input several parameters. In order to solve this limitation, a new way for wave-vortex separation is proposed by using the Linear Stochastic Estimation (LSE). This method has been previously used to extract the conditional coherent structures embedded in turbulent flow. The interest of such method relies on the fact that it is possible to estimate a flow variable which is statistically associated with another variable that drastically reduces the need of input parameters.

A specific designed field experiment has been conducted: a regular monochromatic wave is generated and interacts with some large scale flow structures developing in the wake of the wavemaker rigid structure. Particle Image Velocimetry (PIV) measurements are performed to access the instantaneous velocity field in a vertical plane. The temporal signal of the surface water waves is also simultaneously determined thanks to surface probe measurement. Two test flow configurations are studied by considering monochromatic wave propagating at the same frequency with or against a 0.8m/s flow. In each wave configuration, LSE is implemented to instantaneously estimate the fluctuating wave component  from the raw available fluctuating velocity measurements  by exploiting the correlation between wave temporal signal and instantaneous velocity fields. This procedure is applied to each velocity signal available in each PIV mesh-grid point by accounting for each, the distance from the water surface.Based on this instantaneous flow decomposition, a spectral analysis of the wave velocity field is first conducted.  Whatever the wave propagation direction, some similar results are observed. Even if a wave frequency peak is mainly observed at the water surface wave frequency, some other frequency peaks emerge not only at twice that frequency but also at frequencies that correspond to some combinations of the free surface wave frequency and the frequency passage of the large scale wake flow structures. This clearly indicates that waves act in the whole spectral domain. Second, by analysing the extracted wave velocity field, the wave contribution in the turbulence kinetic energy of the wake flow is higher when waves propagate against the current. Finally, present results are compared to previous wave-turbulence decomposition methods, confirming that waves tend to reduce the vertical mixing by reducing the kinetic energy vertical transfer in the present flow configuration for which the mean velocity vertical  gradient is negative. Present analysis provides a better understanding of the ocean vertical mixing and its associated energy transfer. These results are of great interest for marine current turbine behavioral study (in term of loads, performance and regulation law) for which a precise knowledge of the encountered functioning flow conditions is required.

Presenting Author: Philippe Druault Sorbonne Université

Presenting Author Biography: Philippe Druault is currently Associate Professor at Sorbonne University. He holds a PhD from Poitiers University.
His current research areas include turbulence and tidal turbine renewable energy.

Authors:

Philippe Druault Sorbonne Université
Benoît Gaurier IFREMER
Grégory Germain IFREMER