Friday 11 October 2013

Why not use wavemakers?

"If you are trying to make a good wavemaker, why not use wavemakers?"

If we take the statement 'a good wave absorber is a good wave maker' to its logical conclusion, then why not consider the wavemakers in our wave tanks? Most tanks for simulating deep water waves have a series of bottom-hinged flaps, mounted on the edge of the tank.

How do flap wavemakers generate waves?

Flap wavemakers are suited to generating deep water waves, as their horizontal position profile is not too far away from that of a deep water wave: (see diagram below)

These flaps have water on the tank side, and air on the other, so when they move, their motions only disturb water on the tank side. In wave power parlance, we'd refer to a bank of such paddles as 'terminators': none of the energy in a wave travelling towards these paddles is transmitted beyond them. We also think of these as terminators because of the nominal wave direction with respect to the wavemakers: the wave fronts are nominally parallel to the bank of paddles (waves travel perpendicularly to bank of paddles). Although each paddle operated individually can only generate a wave in one direction, when operated together, the bank of paddles can generate waves at a range of angles. This is done by driving each paddle slightly out of phase with its neighbour.

A standard public demonstration at the Edinburgh wave tank is to generate a regular wave front at 20° to the bank of paddles, then one at -20°, and then to combine these two waves, producing a pleasing quilted surface. Less visually impressive, but of more practical use, is to generate a short-crested spectrum typical of those measured in deep waters. If such a system is good at making typical ocean waves in tanks, surely it should be good for absorbing waves? Actually, we already have the answer to this postulation.

How do absorbing wavemakers absorb waves?

Many experimental wave tanks use force feedback wavemakers, which do absorb waves. When the Edinburgh Wave Power Group developed their wavemakers for their (now ex-) wide tank, force feedback was initially introduced to improve the quality of the waves. Negative feedback ensured that the generated wave was close to the demanded wave, whatever the instantaneous water level on the paddle due to incoming waves radiated by the model or reflected by the facing walls.

How this works is the wavemaking software calculates what force should be experienced on each paddle to generate the desired wave. This is compared to the measured force. The force actually applied works to reduce the difference between the desired and measured force. So during operation the paddles are doing two jobs: they are generating the desired waves, and they are also absorbing the undesired (incoming) waves. At the end of a test run, the demanded force is zero. If the wavemakers are left on, they will continue to absorb waves, acting as 'active beaches'. Active wavemakers are useful when tank testing, as they result in fast 'settling times' between test runs. Tanks without absorbing wavemakers take longer for waves to die down between runs.

One of the reasons why the absorbing wavemakers by Edinburgh Designs operate well as wave absorbers over a wide range of frequencies is that the feedback force they use to absorb the waves is proportional not only to velocity, but has components proportional to mass and spring. A control engineer might call this PID (proportional integral differential) feedback; in wave power this is referred to as 'reactive control'. This means that sometimes the motion of the wavemakers is resisted (external force applied in opposite direction to wavemaker velocity), and sometimes their motion is spurred on (external force applied in the same direction as wavemaker velocity).

What would be needed to do to turn a wavetank into a WEC? 


  • Scale it up: As a rough indication, if the paddles at ECN (see figure above) are good for 10th scale tests of the waves we are targeting, the flaps will need to be 10 times bigger. Perhaps it is more economic to be more inefficient in bigger waves (load shedding) – in that case, use smaller paddles (say 3 times bigger than the ECN paddles?).
  • Turn the tank inside out: line 3 sides of an oil-rig with flap wave makers, ensuring that a significant potion of the directional resource would be captured. Rounded corners of the bank of paddles could be beneficial.
  • Stationary panels below the paddles may be required if the paddles do not extend to the depth of the target waves – this will increase the terminator effect. On the other hand, allowing bigger waves to pass underneath could be a useful load-shedding mechanism.
  • Reconsider the scaled-up design: the implementation of PTO and physical spring might need to be reconsidered due to the size change.
  • Keep the force-feedback: it has been tried and tested in tanks for decades, and will allow absorption from short-crested seas from a range of incidences.
  • Reconsider reactive control: In wavemakers, inefficiency due to reactive control is not an issue, as the aim is to cancel the incoming wave rather than extract power from it. When capturing wave power however, the power gains of reactive control must be weighed up against the losses due to inefficient power chains. For cost effectiveness, it might be necessary to reduce the reliance on reactive control: this could be done by changing the design of the paddles, or by controlling paddle characteristics during operation, for example physical spring, or immersion depth.
  • Find a way of resisting loads in high energy sea states. One option is to raise the paddles above the splash-zone. Such a mechanism would in any case be useful for accessing the paddles for maintenance (as is done with Marine Current Turbines; see below), for adjusting to changing tidal levels, and for adjusting paddle immersion for control. Another consideration is the position of stationary panels below the wavemakers.


Comments from Edinburgh Designs

We have been asked about this in the past by people who have seen our tanks working, and the two main reasons as I see it why there are no wave power devices like this are, firstly economic (the design of the machine is actually fairly complicated compared to many of the existing wave power devices, and there is probably a lack of sites for such machines) and secondly survivability (as you say in your last point) - the machines can only absorb waves up to a certain size before they run out of motion.

In fact the machine we are building at the moment in Edinburgh (FloWave) will be doing exactly as you suggest as it is circular and so will need to absorb all of the waves it generates (there is no wave absorbing beach as in other tanks); it will even re-generate the absorbed wave energy onto the grid.  However we will need to adjust our software to prevent the wave machine from generating any waves that would be too big for the paddles on the opposite side of the tank to absorb, as this would potentially damage the machine.  We can do this because we have full control of the environment, which is clearly not possible in the open sea!

As for the FloWave tank putting energy back into the grid - this is primarily for practical reasons, to remove the need for the large resistor banks that would otherwise be required to dump the excess energy as heat. The machine will, of course, be a net energy consumer (it's no perpetual energy machine!) 

Image credits:

'Imperial College London wide tank' by Edinburgh Designs
'wave base' from wikipedia commons:
'48m x 3m deep paddles at Ecole Centrale de Nantes' by Edinburgh Designs
'Seaflow', MCT's single turbine prototype, from


I am grateful to Aurélien Babarit who, at EWTEC 2013, posed the question that is the title of this blog, and suggested I should write a post on this subject. Thank you to Ian Jason from Edinburgh Designs for his discussion of this idea.

No comments:

Post a Comment


Note: only a member of this blog may post a comment.