Numerical Model Assessment of Tidal Stream Energy Resources in the Severn Estuary, UK

Abstract

Tidal stream systems make use of the kinetic energy of tidal movement to power turbines, in a similar manner to the way in which windmills extract energy from the wind. This method of renewable energy generation is gaining in popularity because of the high predictability of tides, the lower investment needed for tidal turbines, and the lower ecological impact, as compared with other schemes involving barrages or lagoons. It is generally considered that a mean spring peak tidal current of at least 2 m/s is required for tidal stream power to be worth exploiting. In the Severn Estuary, the peak tidal current exceeds 2 m/s, with a corresponding minimum depth of greater than 20 m, and it is thereby a potential location for tidal stream power. Previous studies cannot provide the detailed and precise distribution of tidal stream power in this estuary, and it has therefore been deemed appropriate to undertake a numerical model assessment of tidal stream energy resources in the Severn Estuary. In the present study, an existing finite volume numerical model has been refined, with the inclusion of an algorithm for computing the power density and its mean value across the estuary. The refined model has also been validated against (a) measured tidal currents at four sites, using the method of harmonic analysis; (b) measured tide level hydrographs at five tide gauging stations, taken over a 15-day period and covering the full spring—neap cycle; and (c) in situ velocity measurements at two sites. Finally, the validated model has been used to assess the potential tidal stream energy resources without and with the Severn Barrage, including the distributions of the mean power density over a spring—neap cycle in the Severn Estuary and a detailed assessment of the tidal stream energy resources at two sites near the coast of South Wales. The model predictions indicate that at two sites, the annual power output with the presence of barrage could be reduced by 70—80 per cent, as compared with the value without any structure.