Page 305 - Fundamentals of Ocean Renewable Energy Generating Electricity From The Sea

290 Fundamentals of Ocean Renewable Energy

levels of energy harvesting. In general, the IEC Technical Specification for
Tidal Energy Resource Assessment and Characterization [31] recommends that
energy extraction be accounted for in ocean models when the installed capacity
is greater than 10 MW, or the proposed level of energy extraction exceeds 2% of
the theoretical tidal energy resource.
Energy extraction can be incorporated in ocean models by modifying the
governing equations of flow, and so the turbine is represented in the model
by including the mechanics of the energy extraction process, rather than the
physical structure of the turbine itself [33]. At the scales of relevance to ocean
models, generally additional friction, or a momentum sink approach is used to
represent turbines, but the actuator disc model is also gaining popularity [33].
Actuator disc theory was explained in detail in Section 3.13.
In an early paper on tidal energy extraction, Bryden and Couch [34]
presented a relatively simple 1D case of energy extraction in a rectangular cross-
section channel, based on an additional bed roughness term. Their finding that
10% of the energy flux can be extracted before there is a significant change in the
flow characteristics, remained a popular ‘rule-of-thumb’ for the limits of energy
extraction for many years. Neill et al. [35] applied a similar technique of an
additional quadratic friction term in the momentum equation to a 1D study of
the Severn Estuary/Bristol Channel under various extraction scenarios, finding
that a relatively small amount of energy extraction (particularly at critical points,
relating to tidal symmetry) can lead to a significant impact on the sediment
dynamics of a channel, including far-field effects. However, in common with
Bryden and Couch [34], their 1D model did not account for lateral effects, and
both studies could, perhaps, have overstated the effect of energy extraction.
If turbines occupy only a fraction of the cross-section, the available power
5
drops below that obtainable from a complete fence, since energy is lost in
downstream merging of the wake and the surrounding fluid [36]. For example,
in a 2D model the flow can also by-pass the turbine, and so efficiency of energy
extraction will reduce significantly. Other research makes use of a momentum
sink approach in either 2D (depth-averaged) [37] or 3D [38] models, but it
is the use of 3D models that is of particular interest. With increasing access
to supercomputing facilities and increased power of desktop computers, it is
becoming increasingly common for resource characterization studies to make
use of 3D models [39]. There is on-going research about whether a turbine (or
an array of turbines) needs to be parameterized in these 3D models at the actual
depth of energy extraction (i.e. over the swept area of the rotor), or whether a
simpler ‘depth-averaged’ turbine representation is sufficient. It is important to
note in such discussions that the support structure of the turbine will, in the
majority of device designs, extend over a greater portion of the water column

5. Rather than installing devices individually within arrays, a tidal fence consists of a ‘continuous’
row of turbines, extending across the width of a channel.