Empirical modelling of turbulence and wave-current interaction in tidal streams
Thesis or dissertation
- © 2017 Ross Jennings. All rights reserved. No part of this publication may be reproduced without the written permission of the copyright holder.
The successful development of the tidal stream power industry fundamentally relies on a thorough, quantitative understanding of the available resource. Numerical simulations and laboratory flume experiments have demonstrated that increased turbulence and wave-induced motion can have detrimental effects on the fatigue and performance of prototype tidal stream turbines (TSTs). Knowledge of the relationships between mean current velocity, turbulence and surface waves is limited and presents a significant research gap. This research makes a significant contribution to the field by developing empirical models from in situ data collected within the Humber Estuary. These models estimate the turbulence strength and intensity at a point and through depth given a mean current velocity.
An 18-day deployment of bed-mounted directional wave recorders (DWR) at Foul Holme Spit simultaneously recorded two-dimensional flow velocities and surface wave parameters. Static, vessel-mounted, acoustic Doppler current profiler (ADCP) surveys recorded turbulence through depth near St. Andrews Dock.
The analyses revealed distinct relationships between the mean current velocity, turbulence strength and turbulence intensity at a point which are comparable to recently published results. The inter-tidal relationship between streamwise mean current velocity and turbulence strength is modelled at a point using power regression where α is 0.13 and β is 0.72 with an R² value of 0.8721. The inter-tidal relationship between streamwise mean current velocity and turbulence intensity is modelled at a point using power regression where γ is 14.315 and δ is -0.2316 with an R² value of 0.5482. A newly defined empirical relationship between depth-averaged mean current velocity and turbulence intensity is modelled using power regression where ε is 17.75 and ζ is -0.94 with an R² value of 0.7912. The models derived at a point are tested on the data collected through depth and exhibited strong predictive capability within the order of 0.1 ms-1. The exponential approximation of wave-induced velocity, proposed by Soulsby (2006), was tested and shown to be inappropriate for estimating wave-induced velocities at this scale. A comparative spectral analysis between DWR sample bursts determined that spikes in the turbulence spectra can be attributed to surface wave parameters, thus validating the conceptual model proposed by Soulsby and Humphrey (1990).
- Department of Geography, The University of Hull
- Hardisty, Jack, 1955-; Ferrier, Graham
- Sponsor (Organisation)
- Centre for Adaptive Science and Sustainability
- Qualification level
- Qualification name
- 8 MB