The development of Oyster – A shallow water surging

wave energy converter

Trevor Whittaker1,2

, David Collier2, Matt Folley

1, Max Osterried

1, Alan Henry

1,

Michael Crowley2

1Wave Energy Research Group, School of Planning Architecture and Civil Engineering,

Queen`s University Belfast, Belfast BT9 5AG, UK

E-mail: t.whittaker@qub.ac.uk

2Aquamarine Power Limited,

10 Saint Andrew Square, Edinburgh EH2 2AF, UK

E-mail: dcollier@aquamarinepower.com

Abstract

In 2005 Aquamarine Power Ltd. was formed to

develop Oyster, a near shore flap which is hinge

connected to the sea bed. With a combination of

private equity and grant aid a 350kW Oyster module

has been designed and it is planned to install a

prototype module at the EMEC test site in Orkney

when the nearshore test berth is available. In this

version of Oyster high pressure sea water will be

pumped ashore to drive a Pelton wheel. Ultimately

it is envisaged that Oyster units will be arranged in

clusters feeding power to a power take off unit of

between 3.5 and 5 MW capacity. Arrays of clusters

will form power stations of 20 to 100 MW capacity.

An extensive research and development

programme has produced a very efficient structural

form, which gives Oyster one of the highest power

to weight ratios of all current technologies

combined with high capture factors in the most

commonly occurring seas. The sea bed foundations

and installation technique developed enables Oyster

to be easily removed and reinstalled for major

maintenance when required. This is a feature

normally associated with moored devices.

Although there are other bottom-hinged flap

devices, Oyster is different in several ways and

occupies a different part of the design space. For

example, unlike the other systems it completely

penetrates the water column from the water surface

to the sea bed. Although it might be considered that

such a system would be vulnerable in extreme seas,

extensive wave tank modelling has shown that the

flap intrinsically decouples from the wave as the

oscillation increases and that the wave loads

experienced are manageable in the three operational

modes; generating, undamped and parked on the sea

bed. However, model tests show that Oyster can

remain generating in all sea-states including

plunging breakers.

This paper charts the evolution of Oyster

presenting some of the research that has led to the

current design. An outline of the impending sea

trials of a prototype demonstration unit is given

along with the projected outcomes.

Keywords: Oyster, shallow water, flap, surging,

loads, performance

Introduction The research group at Queen’s University Belfast

(QUB) has been actively involved in wave power

R&D since the mid 1970’s. Most importantly they

co-ordinated the design, construction and operation

of two full-scale prototype shoreline oscillating

water column (OWC) devices on the Isle of Islay.

The first of these was a 75kW machine completed in

1990 and operated for research purposes until 1999

when it was decommissioned. The second device,

LIMPET, [1], has a 500kW installed capacity and is

currently owned and operated by Wavegen. The

significance of the group’s work was recognised by

the award of The Royal Society ESSO Energy Prize

in 1994.

Experience gained from the Islay prototypes

indicated to the group that OWC’s are structurally

inefficient because of the need to enclose the

working surface (excluding cases where the

structural function can be shared e.g. breakwaters,

or can be formed cheaply e.g. by tunnelling in rock).

Consequently other solutions were sought that

substantially reduced the structural content of the

primary part of the converter. The most promising

solution considered resulted from the fact that in

water depths of around 10m the horizontal fluid

particle motion is amplified compared to that in

deeper water and that this could be exploited by

surging devices.

1. Background research In 2002, with funding from the UK Engineering

and Physical Science Research Council, the team

began researching a group of devices which they

described as oscillating wave surge converters

(OWSC’s). The mechanically simplest form of this

family of devices is hinged flaps, with the hinge

either at the top or bottom. Initially only shoreline

configurations were considered and testing was

conducted in 2D in the 0.33m wide wave tank at

Queen’s. This led to the testing of nearshore

configurations and subsequently to 3D tests in a

4.5m wide wave tank. A 1:20 seabed slope similar

to that experienced off Islay was modelled at 1:40th

scale and a set of Bretschneider seas programmed

and weighted to be representative of the inshore

wave resource in the North Atlantic off the Western

Isles off Scotland. Top-hinged flaps located in a

recess with an inclined rear wall were initially tested

as they were considered the most logical evolution

from LIMPET, with the swinging flap and hydraulic

power take off replacing the OWC structure and air

turbine. The principal hydrodynamic shapes and the

models used are shown in figure 1.

Shoreline Nearshore

Figure 1 Configurations tested

Benchmarking experiments were performed, as

described by Folley et al [2], in which the OWSC

was compared to LIMPET, [3], and the Japanese

Pendulor device, [4]. These showed that the flap

devices were substantially better than the inclined

water column in the shorter period lower energy

seas while the reverse was true in the longer period

high energy seas. This demonstrated one of the most

desirable features of flap devices, their natural

tendency to increase capture efficiency in the

smaller, most commonly occurring, seas. Compared

to OWC’s with the opposite characteristic, this

enables the installation of a smaller capacity PTO

system on flaps which runs at a higher load factor

and this natural decoupling characteristic aids

survivability in the large high energy seas.

Although shoreline mounted top-hinged flaps are

a potentially viable system, it was decided to

concentrate on the nearshore configuration because

of its greater commercial potential. This

configuration evolved into a simple bottom-hinged

flap fastened to the seabed without any surrounding

structure and named OysterTM.

2. Description of OysterTM

In its present configuration OysterTM is

essentially a wave powered hydroelectric plant

located at a nominal water depth of 12m which in

many locations is relatively close to the shoreline.

The system comprises a buoyant flap, 18m wide and

10m high, hinged at its base to a sub-frame which is

pinned to the sea bed using tensioned anchors. The

surge component in the waves forces the flap to

oscillate which in turn compresses and extends two

hydraulic cylinders mounted between the flap and

the sub-frame which pumps water at high pressure

through a pipeline back to the beach On the shore is

a modified hydro-electric plant consisting of a

Pelton wheel turbine driving a variable speed

electrical generator coupled to a flywheel. Power

flow is regulated using a combination of hydraulic

accumulators, an adjustable spear valve, a flywheel

in the mechanical power train and rectification and

inversion of the electrical output. An outline

schematic of the 350kW unit to be installed at the

EMEC test site off Orkney is shown in figure 2.

Figure 2 General arrangement of OysterTM

3. Hydrodynamic performance The hydrodynamic performance of Oyster has

been assessed and enhanced using a combination of

analytical, numerical and experimental modelling.

This has formed an integral part of the design loop

as results from model tests have been used to

develop the analytical theory of how Oyster works,

which in turn has been used to identify the

appropriate model testing to be performed. Wave-

tank modelling of Oyster was initially performed

using a 1/40th scale model, whilst the most recent

testing has been done at 1/20th scale. All testing has

been carried out in the wave-tank at Queen's

University Belfast, which has 6 sector-carrier wave

makers capable of generating complex and realistic

sea-states. Although some monochromatic testing

has been undertaken, the majority of the testing has

used sea-states with a Bretschneider spectrum.

A time-domain numerical model of the Oyster

hydrodynamics and PTO system has been developed

using a second-order differential equation of the

hydrodynamics and calibrated using results from the

wave-tank model. Even though for computation

speed and model simplicity the numerical model

lacks the convolution integral or additional modes

typically seen in a hydrodynamic model, it has

typically predicted the average power capture of

Oyster to within 10% of that obtained with the

wave-tank experimentation. This is related to the

relatively small magnitude of the wave radiation

forces.

The modelling undertaken indicates that Oyster

operates by coupling strongly with the incident

waves, without being highly tuned; that is, there is a

large wave torque and the natural pitching

frequency is not close to the predominant incident

wave frequencies. It is shown elsewhere, Folley et

al. [5], that the coincidence of natural tuning and

maximised wave torque are largely mutually

exclusive for this type of wave energy converter.

However, in shallow water the benefits of tuning are

limited due to motion constraints and viscous losses

at large amplitudes of motion, though it is still

beneficial to maximise the flap pitch stiffness with

excess buoyancy to keep the natural pitch frequency

as close as economically possible to the most

common incident wave frequencies.

The wave force experienced by Oyster is related

to the change in horizontal acceleration of the water

particle motions and thus tends to increase in

shallow water due to shoaling. For a 10 second

wave in a water depth of 12 metres, this is

equivalent to an increase of 50% in the wave force

experienced relative to the force experienced in

deep water Folley et al. [6]. Maximum interaction

with the horizontal acceleration of the water

particles and hence wave force occurs when the flap

and substructure penetrates the full depth from sea

bed to surface. For such a flap that is only a fraction

of a wavelength wide the wave force increases

approximately as the square of the flap width

resulting in a relatively wide flap. However,

maximum width is limited by the reduction in

performance due to wave force phase incoherence

across the flap, together with structural and

installation considerations. Figure 3 shows the

capture factor, defined as the percentage of the flap

width from which 100% of the incident energy is

extracted, in a variety of sea states.

Figure 3 Capture factor vs sea state & flap

width

These results show that the capture factor is

largest with the smallest sea-states, short energy

periods and small significant wave heights, which

help to even out the power capture. Indeed results

from the numerical simulation indicate that the

power capture is almost independent of wave energy

period and depends primarily on significant wave

height, though it is currently unclear if this is due to

an underlying characteristic, or whether it is a mere

coincidence.

The average power capture of the device

deployed adjacent to a typical North Atlantic coast

was calculated by taking the performance in each

sea multiplied by an occurrence weighting factor.

This was determined by weighting each sea to give

an annual average exploitable incident wave energy

of 19 kW/m. This produces an average power output

at the hydraulic cylinders of around 200kW. A

much more extensive analysis based on 26,500 seas

calculated from 19 years of hindcast wave data at

the EMEC test site produced a similar performance

figure.

As with any other wave power device, reducing

losses such as vortex shedding losses around the

edges increases the power output of Oyster. This

can be achieved by either thickening the flap as a

whole or just the edges. Though an increase in

power capture was found in both the numerical and

wave-tank models, the increase was relatively

modest and could not be justified for a prototype; it

is possible that future production models will have a

more complex, optimised, shape.

The Oyster wave energy converter is directly

connected to the seabed in a water depth of

approximately 12 metres so that tidal level would be

expected to have an effect on the hydrodynamics

and power capture. At high tide the Oyster flap will

be more submerged and the waves overtop the flap

more easily reducing the wave force and therefore

the power capture. At low tide the maximum

amplitude of motion of the centre of wave pressure

is effectively reduced because it is closer to the

flap’s axis of rotation, which reduces the power

capture in the more energetic sea-states. That power

capture typically reduces at both high and low tide

is expected since the freeboard and design of the

flap has been optimised for the mean water level

(MWL); however, the reduction in power capture at

different tide levels is of importance.

Taking the tidal variation for the EMEC wave

energy test site as an example, although the stated

tidal range is approximately 3.0 metres, for 90% of

the time the water level is less than 1.0 metres from

the mean. The effect of water depth has been

investigated using the wave-tank model and figure 4

shows the average power capture for the six

standard sea-states at 1.0m above MWL, MWL and

1.0m below MWL obtained from these tests. The

reduction in power capture at high and low tide is

typically less than 12%. Thus, whilst there is a small

reduction in power capture with tidal variation, the

Oyster wave energy converter is relatively

unaffected by typical variations in tidal level.

Figure 4 Power capture variation vs tide

The directional variation and spread of the

incident waves will also influence the power capture

of Oyster because it couples with the surge

component of the waves, which have an azimuthal

orientation and the wave force reduces when the

wave’s crest is not parallel to the flap’s face. For a

small body it would be expected that the wave force

will reduce in proportion to the cosine of the angle

between the direction of wave propagation and the

orientation of Oyster’s flap, which would typically

result in a reduction in power capture proportional

to the cosine squared. As the Oyster flap gets wider

the reduction in wave force will be greater due to

incoherence of the wave acting on the flap. WAMIT

has been used to investigate the reduction in wave

0.00.10.20.30.40.50.60.70.8

Cap

ture

fac

tor

0.00.10.20.30.40.50.60.70.8

Cap

ture

fac

tor

+1.0 m MWL

MWL

-1.0 m MWL

force, which shows that for the 18 metre wide

Oyster the reduction in wave force is only slightly

greater than for a small body as shown in figure 5

Figure 5 Wave force WAMIT, vs

orientation

This reduction in wave force and power capture

with angle of incidence has been verified using the

wave-tank model, and figure 6 shows that the

reduction in power capture is approximately

proportional to the Cos of the angle of incidence to

the power 2.1.

Figure 6 Power reduction with orientation

To estimate the effect that the directional spread

of the incident waves will have on power capture a

non-linear wave propagation model is used to

transform the waves from offshore to the proposed

location of the Oyster prototype to be deployed at

the EMEC wave energy test site on the 10 metre

depth contour, (relative to chart datum). Applying

the reduction in power capture derived from the

wave-tank model testing to the EMEC data indicates

that with the optimal orientation of the flap the

power capture is reduced by about 3% due to the

directional variation of the incident waves. This

relatively small reduction in power capture due to

directionality is due to the inherent tendency for the

most energetic waves to come from a relatively

narrow sector orientated in a westerly direction and

because as the water depth decreases the wave

fronts refract to become more parallel to the depth

contours.

4. Wave loading and survivability Survivability is the first and most important

achievement of any wave power system and

consequently a substantial part of the test

programme has been wave load testing. The

objective was to ascertain the envelope of heave,

surge and torque loads on the foundations of Oyster

in both the extreme and the most frequently

occurring seas. This data provided the input to both

the foundation design and the fatigue analysis of the

structure. The tests were conducted in the 4.5m wide

wave tank at Queen’s university using a purpose

built load table on which a 40th scale model of

OysterTM was mounted.

Two distinct sets of experiments were conducted

to meet different objectives. In the first set the

envelope of the maximum loads was measured in a

range of seas up to the most extreme seas which

exist at a water depth of up to 12.4m. In the second

set the most commonly occurring seas were used in

order to provide data for the fatigue analysis of the

structure.

In OysterTM the torque transmitted from the flap

to the foundations is limited by the torque applied

by the PTO system and there are three operational

modes to consider; the flap parked on the sea bed,

free undamped oscillation and generating. As

expected the highest vertical loads were experienced

with the flap parked on the sea bed due to the

pressure produced by the plunging breakers on the

surface above. As these are acting downwards they

are easy to accommodate. Undamped oscillation

resulted in the lowest surge loads. However,

although the surge loads are 50% higher in the

generating mode, the foundations are designed to

withstand these to allow generation during extreme

seas. The coincidence of heave and surge loads

whilst generating in an extreme sea is given in

figure 7.

Figure 7 Coincidence of heave and surge

loads

One of the perceived problems with bottom

hinged flaps is the possibility of them hitting the sea

bed when undamped. This was tested by focussing

the seas on the model to produce an extreme

plunging breaker. After repeated tests with different

sea-states the maximum oscillation observed was

less than 800 from the vertical. Excess buoyancy in

the flap, decoupling from the wave as the oscillation

angle increases and a squeeze film effect between

the flap and the sea bed limits the angular excursion.

Figure 8 shows the model about to be hit by a

plunging breaker and figure 9 the effect on rotation.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10 20 30 40 50 60 70 80 90

No

rmal

ise

d w

ave

torq

ue

Oscillator angle (degrees)

7

9

11

13

Cos

0

50

100

150

200

250

300

350

0 20 40 60 80

Po

we

r ca

ptu

red

(kW

)

Flap angle (degrees)

18m Flap n=2.1

Te = 7 s, Pi = 10 kW/mTe = 9 s, Pi = 20 kW/mTe = 11 s, Pi = 20 kW/mTe = 13 s, Pi = 40 kW/m

Figure 8 Wave loading – extreme seas

Figure 9 Flap rotation after impact

5. Installation and recovery A major cost in marine renewables which is often

underestimated is installation and long term

maintenance. A perceived advantage of moored

devices is that they can be readily disconnected and

towed to port for major maintenance. This is more

difficult with devices attached to the sea bed. A key

development with Oyster has been the design of a

system which enables self installation and removal

with the aid of small service vessels. As the flap and

the sub-frame is an integral unit the flap provides

the buoyancy to control the lowering of the sub-

frame onto the sea bed foundation pads. Once the

sub frame is locked into the clamping mechanisms

the chambers in the flap are flooded sinking it onto

the sea bed thus enabling divers to make the final

connections next to a static object. When the

installation is complete the flap is de-ballasted and

the unit commences generation. The procedure is

reversed for removal in the event of the requirement

for major maintenance.

6. Future prospects The ultimate objective is to build wave power

stations of 20 to 100MW. Oyster is a modular

system with the individual flaps arranged in clusters

feeding power to a single PTO system of between 3

and 5MW capacity. Research is underway to

determine the hydrodynamic performance of

clusters of flaps in different geometric patterns. In

addition during the next 12 months work will be

ongoing modelling the PTO system with multiple

hydraulic inputs to a single Pelton wheel and

generator. With device clusters it is expected that

phase variation between the units will result in a

smoother power output as well as economies on

foundations due to phase cancellation of loads.

The first Oyster demonstrator will provide

essential information on cost and performance

which will enable the development of a techno-

economic model to assist the design process in

future. This will permit the detailed engineering

design to evolve based on ‘real sea’ experience and

the ‘real cost’ of components. During the first 12

months of the sea trials five stages of achievement

have been identified;

Installation and survival

Conversion of sea power to a hydraulic output

Raw electrical power output

System control to improve power quality

Detailed design of a cluster

A further aspect of the future development

programme will be wave powered desalination by

feeding the high pressure water directly to reverse

osmosis tubes.

7. Concluding remarks

The Oyster wave power system is about to be

demonstrated after an extensive research and

development programme during the last six years. It

is a wave powered hydroelectric plant in which a

nearshore bottom hinged flap pumps water ashore to

drive a Pelton wheel coupled to a generator. The

most significant aspects are;

Exploitation of the amplified surge component in

shallow to intermediate depth water

Highest capture efficiency in most commonly

occurring seas with a full depth flap from sea bed to

surface.

Continued generation in extreme seas without

excessive PTO capacity due to the flap’s natural

characteristic of decoupling as the angle of rotation

increases.

One of the highest power to structural weight

ratios relative to other systems under development.

The development of an installation and removal

system which can be undertaken in short weather

windows without specialist vessels.

A modular system in which individual flaps can

be arranged in clusters pumping water to a single

PTO system of up to 5MW capacity. The clusters

can be combined in 100MW arrays.

Acknowledgements

The work described has been funded by

Aquamarine Power Ltd, The Department of Trade

and Industry in the UK, The Scottish Executive and

The Engineering and Physical Science Research

Council.

References

[1] T. Heath, T.J.T. Whittaker, and C.B.

Boake. The Design, Construction and

Operation of the LIMPET Wave Energy

Converter (Islay, Scotland). in 4th European

Wave Energy Conference. Aalborg, Denmark,

2000

[2] M. Folley, T. Whittaker, and M. Osterried.

The Oscillating Wave Surge Converter. in

Fourteenth International Offshore and Polar

Engineering Conference. Toulon, France, 2004

[3] T. Whittaker, W.C. Beattie, M. Folley, C.

Boake, M. Wright, and M. Osterried.

Performance of the LIMPET wave power plant

- prediction, measurement and potential. in 5th

European Wave Energy Conference. Cork,

Ireland, 2003

[4] T. Watabe, H. Yokouchi, S. Gunawardane,

B. Obeyesekera, and U. Dissanayake.

Preliminary study of wave energy utilisation in

Sri Lanka. in ISOPE. Stavanger, Norway, 2001

[5] M. Folley, T. Whittaker, and J. van't Hoff.

The design of small seabed-mounted bottom-

hinged wave energy converters. in 5th

European Wave and Tidal Energy Conference.

Porto, Portugal, 2007

[6] M. Folley, T.J.T. Whittaker, and A. Henry,

The effect of water depth on the performance

of a small surging wave energy converter.

Ocean Engineering, 34(8-9): p. 1265-1275,

2007