The Owel Wave Energy Converter
Author: John Kemp, Extra Master, PhD, FRIN, FNI OWEL, UK
 | SYNOPSIS |
The genesis and the principle of the OWEL wave energy converter are described.
The results of a DTI SMART sponsored feasibility study, including a mathematical modelling investigation and a towing-tank test programme, are presented and discussed. A planned development phase, involving the design, construction and testing of a large-scale model is outlined, and some preliminary consideration is given to the commercial potential of full-scale wave energy converters.
| 1. INTRODUCTION |
Offshore Wave energy Ltd (OWEL) was formed in early 2001 to investigate and develop a new concept for converting ocean wave energy into useful power. Considerable preliminary work had been carried out, over a number of years, by the author, and this led to the formation of a virtual company which included Sycamore Innovation Management, IT Power and Business Link Wessex.
A proposal for a feasibility study resulted in a DTI SMART Award in March 2001, and the virtual company was then converted into a formal limited company to conduct the feasibility study and to facilitate further development.
| 2. PRINCIPLE OF OPERATION |
2.1 Design Criteria
Following a review of existing and proposed systems for harnessing wave energy, the OWEL wave energy converter (WEC) was designed with a number of criteria in mind. These included:
1) Bearing in mind the rapid decrease in wave energy in shallow water, the WEC should be sited offshore, preferably in water deeper than 40 metres.
2) A floating unit should be used to facilitate offshore deployment and to ensure that performance is independent of tidal range.
3) A robust structure with low maintenance is required for operation in energetic wave spectra. There should be no moving parts in contact with the water and the structure should be simple and inexpensive to construct.
4) The horizontal component of wave motion should be used since wavelengths are many times greater than wave heights. Also, the horizontal motion is uni-directional whereas the vertical motion is oscillatory.
5) The low-grade wave energy should be amplified to produce higher air pressure for supply to a turbine.
6) Floating units should be designed to absorb only a small proportion of the energy in storm waves.
2.2 Implementation
With the above criteria in mind, a WEC unit was designed in the general form of a horizontal duct floating such that the freeboard would be approximately equal to the amplitude of the ambient waves. A unit has a length equal to the longest wave in which it is intended to operate and this feature, together with a small waterplane area, ensures that a platform formed from multiple units would remain steady in a seaway.
The duct in each unit is open at one end, and the mooring system of the platform ensures that this end is presented to the incoming waves. The duct has angled side-plates and an angled bottom-plate or ramp. The air in the trough behind an incoming wave is trapped as the following wave seals against the top of the duct.
This air is then compressed by the angled side-plates and delivered to a compression manifold. An air take-off, through a one-way valve, feeds the compressed air to a reservoir. Adjacent to the manifold is a baffle system which disperses the remnant energy in spent waves so that they do not reflect back along the duct to interfere with following waves. The air collected in the reservoir may be used to drive a compressed air turbine and thence generate electricity.
Figure 1. Schematic of WEC unit principle |
2.3 Dimensions
As part of the feasibility study, a tank-testing programme has been completed using a model 1.8 metres in length, 0.6 metres in breadth and 0.4 metres in depth.
It is proposed that the feasibility study should be followed by a development phase in which we would seek to design, construct and test a larger-scale model some 20 metres in length, 3 metres in depth and 4 metres in breadth.
A full-scale device would need to be longer than the longest wave in which it would be designed to operate. Practical dimensions for a unit might be 200 metres in length, 33 metres in breadth (narrowing to 7 metres), and 30 metres in depth, with 6 such units rigidly disposed side-by-side to form a roughly triangular platform 200 metres long and 200 metres in width (narrowing to 40 metres).
It is estimated that such a platform could produce greater than 6 megawatts of useful power when optimised for the wave conditions.
Figure 2. Platform Schematic |
| 3. FEASIBILITY STUDY |
The feasibility study comprised three main components. A mathematical modelling study, a tank-testing programme and an assessment of the commercial potential of the WEC design. The mathematical modelling and the tank-testing components were conducted for OWEL by teams from the QinetiQ organisation. The results of the three components are presented and discussed in the following sections of this report.
3.1 TANK-TESTING PROGRAMME
3.1.1 Purpose The purpose of the tank-testing programme was mainly to obtain experimental data on the performance of a model WEC for confirming and calibrating the mathematical modelling predictions. This was to give added confidence to the mathematical results and, particularly, to the estimates of power-output from full-scale WECs.
3.1.2 Management The supervision of the tank-testing programme, including the production of a report, was put out to tender and responses were received from the QinetiQ Organisation, the University of Southampton's Wolfson unit, and the Southampton Institute for Further Education.
The QinetiQ bid was accepted on the basis of price and perceived capability. However, in order to keep costs to a minimum, and for convenience of access, it was decided to use the ship towing-tank facility at Southampton Institute. The towing-tank hire was the subject of a separate contract between OWEL and the Institute.
3.1.3 Experimental Design The overall design of the tank-testing programme was specified by
OWEL, while the detailed implementation of the programme was closely and efficiently supervised by the QinetiQ team.
The work programme consisted of:
(i) | Discussion with QinetiQ concerning the test conditions. |
(ii) | Installation and calibration of QinetiQ equipment at Southampton Institute. |
(iii) | Testing the specified WEC model configurations under 8 specified wave conditions. |
(iv) | Acquisition and logging of data. |
(v) | Analysis and reporting of results. |
The model was supplied by
OWEL and was a redesigned version of an earlier model which had been the subject of preliminary tank-testing in October 1999. The model dimensions were as noted in Section 2.3 above and for the test programme, it was secured near the centre of the towing tank.
The independent variables in the test-programme were the model configuration and the generated wave spectra. The model was tested in both free-floating (loosely tethered) and clamped modes, and at freeboards of 5 centimetres and 10 centimetres.
Trials were conducted for 3 positions of the side plates, 3 settings of the baffle system, and 3 positions of the ramp.
The model was also tested with no bottom plate fitted. For each of the configurations, the model was subjected to 8 different wave spectra.
The dependent variables in the test programme were mainly the air pressures, P1 and P2 as measured on the supply and delivery sides respectively of the non-return valve. In addition, the motion and accelerations experienced by the model were measured in three dimensions by the QinetiQ "Crossbow" sensor fitted to the deck of the model.
In all, over a test period of 5 days, some 117 test runs were completed. A description of the test conditions and the wave spectra generated, together with the Tank Testing Results, is contained in the resultant QinetiQ report, "Tank Testing of a Wave Energy Converter".
3.1.3.1 Implementation
With the above criteria in mind, a WEC unit was designed in the general form of a horizontal duct floating such that the freeboard would be approximately equal to the amplitude of the ambient waves. A unit has a length equal to the longest wave in which it is intended to operate and this feature, together with a small waterplane area, ensures that a platform formed from multiple units would remain steady in a seaway.
The duct in each unit is open at one end, and the mooring system of the platform ensures that this end is presented to the incoming waves. The duct has angled side-plates and an angled bottom-plate or ramp. The air in the trough behind an incoming wave is trapped as the following wave seals against the top of the duct.
This air is then compressed by the angled side-plates and delivered to a compression manifold. An air take-off, through a one-way valve, feeds the compressed air to a reservoir.
Adjacent to the manifold is a baffle system which disperses the remnant energy in spent waves so that they do not reflect back along the duct to interfere with following waves. The air collected in the reservoir may be used to drive a compressed air turbine and thence generate electricity.
Figure 2a: Artist's impression of the Grampus. |
3.1.4 Results
3.1.4.1 P1 Pressures
As was expected, the pressure, P1, developed in the manifold (the supply side of the non-return valve) of the model exhibited a series of peak values as each successive wave created a pulse of compressed air. The most significant measurement was the mean of these peak values over the duration of a test run.
All pressures are given in the QinetiQ report in millimetres of water, and it was hoped that the P1 mean peaks would, under favourable model configurations, shows pressures greater than the amplitude of the associated waves.
This proved to be the case so that, in 58 of the 117 test-runs, the mean peaks P1 pressure was over twice the wave amplitude and, in 5 cases, the mean peaks P1 pressure was more than three times over the wave amplitude.
This is an important result, showing that it is possible to achieve the multiplication of low-grade wave energy set out as a requirement in section 1.2 of this report.
3.1.4.2 P2 Pressures
The P2 pressures, measured on the air reservoir (delivery) side of the non-return valve should, if the valve had been perfect and the system completely air-tight, have increased exponentially to approach the mean peak P1 values. In fact there was, inevitably, a threshold pressure below which the valve would not open, and there appeared to be some loss of accumulated air pressure on the delivery side of the non-return valve. These two effects simulate the working conditions of a practical WEC.
In such a device, the non-return valve would only open when the P1 pressure exceeded the residual pressure in the reservoir, and there would be a steady loss of pressure between wave pulses due to the outflow of air to the turbine. However, the correspondence was in no way precise and the values measured for P2 can only be taken as a very general indication of what might occur in a practical system. With that proviso, it is encouraging that, in 27 of the test runs, the mean P2 pressure exceeded the wave amplitude.
3.1.4.3 Model Motion
In service, it is envisaged that full-scale WEC platforms would be designed so that they remain as steady as possible in a seaway, with minimal motion in response to the ambient wave conditions. To this end, it is expected that they will be semi-submersible structures with small waterplane areas.
In order to simulate this effect, the test model was clamped into position for the majority of the test runs. However, for 26 runs, it was loosely tethered but, even in this condition, it remained remarkably steady in the water, with a maximum pitch angle of the order of only 2 degrees
3.1.4.4 Effect of Sideplates
Three different positions of the sideplates were tested, inner, middle and outer. There was some indication that the narrower setting was more efficient for shorter wavelengths and that the wider setting was more efficient for longer wavelengths. This was not a pronounced effect, which implies that there would be little penalty in designing a practical WEC with a fixed, mid-position for the sideplates.
3.1.4.5 Ramp Angle
Two different ramp angles were investigated, and also the case of a flat bottom-plate and no ramp fitted. For longer waves, the no-ramp configuration created the higher pressures, with the upper ramp position being the most effective for shorter waves. Again, the variation was not great, which implies that there would be little penalty in designing a practical WEC with a ramp at an intermediate angle.
The model was also tested with the bottom plate completely removed so that, on entering the WEC, the waves retained the characteristics of deep-water, rather than shallow-water, waves. Under this configuration, the air pressures generated were significantly reduced.
3.1.4.6 Baffle Setting
Three different settings of the baffle system were tested, - open, half-open and closed but the results showed little difference in the pressures generated. Again, there would be little penalty in adopting a fixed middle position for the baffle setting in a practical WEC. However, more research is needed into the operation and detailed design of the baffle system.
3.1.4.7 Freeboard
Two freeboard settings were tested, these being 10 centimetres and 5 centimetres respectively. The 5 centimetre freeboard was nearer the amplitudes of the generated waves, and results show higher pressures in this condition. Freeboard appears to be an important factor and it would be desirable to control this to match ambient wave amplitudes in a practical WEC.
3.1.5 Tank-Testing Conclusions.
Pressures generated in the WEC model manifold (P1) were equivalent to a head of water equal to three times the amplitude of the associated waves under favourable model configurations. It is expected that the ratio of air pressure to wave amplitude would increase with increasing scale since the energy in a wave varies as the square of its amplitude.
Pressures generated in the WEC air reservoir (P2) were broadly indicative of what could be expected in an operational device where there would be a steady outflow of compressed air to drive a turbine. Steady P2 air pressures equivalent to a head of water greater than the wave amplitude were observed under favourable model configurations.
There would be little penalty, in terms of device efficiency, in designing a practical WEC with fixed positions for the sideplates, the ramp and the baffle system. This is an important finding since it enables a WEC to meet the robustness criterion, mentioned in section 1.2, that there should be no mechanical moving parts in contact with the sea.
The freeboard of a practical WEC should be controllable so that it can be matched to the mean amplitude of the ambient wave conditions. This could be achieved by means of a simple ballasting system which would not require any moving parts in contact with the sea.
The tank-testing programme, in its own right, has provided results which are encouraging for the future potential of the OWEL WEC concept. However, the main purpose of the programme was to support the mathematical study and to input results which could be used for calibrating the mathematical model. Direct comparisons between the mathematical predictions and the tank-testing results were made and are included in the QinetiQ report "Theoretical analysis of a Wave Energy Converter".
3.2 MATHEMATICAL MODELLING STUDY
3.2.1 Purpose and Design
The mathematical modelling study was planned to be conducted in two phases:
The first phase was to simulate the operation of the physical, tank-testing model in mathematical terms. A set of equations were produced which represented the hydrodynamics and the aerodynamics of the wave/air interactions under the planned tank-testing conditions, such that the output in terms of generated air pressures and power could be estimated. At the end of this phase, the mathematical predictions were compared with the tank-testing results so as to validate and calibrate the mathematical model.
In the second phase, it was intended that the validated mathematical model should be used to predict the output from a full-scale WEC under varying wave conditions, and to throw light on the effects of varying the configuration of the WEC. The full-scale output was successfully predicted in terms of generated air pressures, power production, device efficiency and air flow rate.
The design and implementation of the mathematical study drew on the QinetiQ team's experience of modelling the viscous wave propagation effects in computational fluid dynamics (CFD) and two-phase modelling. This helped to clarify the initial stages of the air compression process.
The analytical model was based on the underlying physics of wave energy conversion. It was developed in the PC-based mathematical environment, Maple.
The following sections discuss the results obtained by the mathematical modelling study and contained in the report by Dr Paul Steer of QinetiQ Ltd, entitled "Theoretical Analysis of a Wave Energy Converter".
3.2.2 Phase 1 Work Programme
The work undertaken by the QinetiQ team in phase 1 of the study was agreed in consultation with OWEL, and included a literature study to identify an appropriate mathematical modelling methodology, and clarification of the physical effects (hydrodynamic and aerodynamic) which needed to be simulated.
The flow processes were modelled and an appropriate simulation program was designed. The simulation was implemented and tested, and then used to predict optimal device configurations and air pressure outputs for comparison with the tank-tests.
3.2.3 Comparison with the Tank-testing Results
The results of phase 1 of the mathematical modelling study were presented in a preliminary report in time for comparison with the tank-testing results. The phase 1 results included predictions of the efficiencies and air pressures generated, as influenced by the effects of changing freeboard, wave height and wavelength.
Although they were not presented in a form which was directly comparable with the results of the tank-testing programme, it was evident that the air pressures measured in the tank-tests exceeded the mathematical model predictions by a factor of ten or more.
3.2.4 Phase 2 - Revision of the Mathematical Model
In the light of the tank-testing results, the theoretical model was revised to improve the accuracy of its predictions. This was mainly a matter of amending the assumptions made concerning the volume of air trapped as each wave trough entered the WEC.
A more realistic model was implemented for the closure of the air space by a wave crest, and the volume of trapped air was inferred from the observed pressure rises in the tank-testing programme. With appropriate calibration factors in place, the mathematical model produced similar results to the tank-tests.
3.2.5 Phase 2 Results
Air pressure predictions from the revised mathematical model gave similar values and showed generally similar trends to those measured in the tank-tests. They confirmed that the WEC freeboard should be adjusted to match the wave amplitude for optimum results. They also confirmed that the angle of the sideplates had little effect on the air reservoir (P2) pressures.
In general, the revised computer model appeared to be capable of predicting the pressures observed in the WEC, but the predicted pressure rises were greater than those measured by from 10% to 100%. This is considered to be good agreement for an idealised model of this type.
3.2.6 Large Scale Performance
3.2.6.1 Scale Factors
Predictions were made for the performance of WECs over a range of scales from the tank-testing model of length 1.8 metres to a full-scale WEC of length 180 metres. In addition to P1 and P2 pressure rises, the efficiency, output power and air flow rates were estimated. The incident waves were scaled so that the length was 0.64 times the internal WEC length, and the wave amplitude was given the somewhat large value of 0.03 of a wavelength.
3.2.6.2 Efficiency
The WEC efficiency is predicted to fall with increasing scale factor, but it is expected that this effect can be reduced or eliminated by appropriate design choices.
3.2.6.3 Power and Air Flow
The output power is predicted to increase dramatically with scale factor, reflecting the greatly increased energy flux of larger incident waves (see fig. 3 below). A power output of 3 megawatts is predicted for a scale factor of 100, which would represent a WEC unit 180 metres long by 60 metres wide in waves scaled as noted in section 3.3.6.1 above.
On this basis, a platform comprising three 60 metre wide units (or six 30 metre wide units) would produce 9 megawatts.
The air flow rate is similarly predicted to increase very rapidly with scale factor.
Figure 3. Power output v. Scale factor |
3.2.7 Mathematical Modelling Conclusions
3.2.7.1 Phase 1
Phase 1 of the study identified a structure and methodology for a mathematical model of the operation of the OWEL WEC. This model was used for predicting the performance of a WEC with the dimensions of the physical model used in the tank-testing programme. In the event, the air pressures measured in the tank tests greatly exceeded those predicted by the mathematical model.
The mathematical model was revised and calibrated in the light of the tank-testing results
3.2.7.2 Phase 2
In phase 2 of the study, the revised model was used to predict the performance of WECs up to 100 times the length of the tank-testing model.
The revised model confirmed the tank-testing conclusion (see section 3.1.4.4) that there would be little penalty in designing a practical WEC with fixed, mid-position, sideplates.
The revised model confirmed the tank-testing conclusion (see section 3.1.4.7) that freeboard is an important parameter and that it would be desirable to control this to match ambient wave amplitudes in a practical WEC.
The predictions relating to efficiency indicated that careful attention should be given to WEC internal configuration and to the design of the compressed air collection system in order that efficiency should not deteriorate with increasing scale.
The output power is predicted to increase rapidly with increasing scale. It is estimated that a 180 metre long WEC would be capable of producing commercial quantities of power of the order of several megawatts.
The air flow rate is also predicted to increase rapidly with increasing scale. This provides the potential for efficient turbine operation.
| 4. DEVELOPMENT PHASE |
Following on from the highly encouraging results of the feasibility study, it is planned to go on to a development phase for the WEC project. This would include further mathematical modelling, leading to the design, construction and testing of a large-scale model.
It is probable that the EEST Companyâs new wave testing facility at Blyth would be used and, with the capacity of this facility in mind, a model of some 20 metres in length is envisaged. It is understood that waves up to a metre in height can be generated, and this would be adequate for the proposed model scale.
As in the case of the feasibility study, the development phase would provide predictions of the output of a full-scale WEC in terms of air pressures generated, volumes of air flow, efficiencies, and power outputs. Compared with the feasibility study, results from the large-scale model would be expected to lead to more precise predictions of the output of full-scale WECs.
In general, the development phase is expected to provide sufficient information to be able to make a positive decision to go on to a final project phase of designing, constructing and deploying a full-scale prototype. | 5. FULL-SCALE WAVE ENERGY CONVERTERS |
5.1 Prototype
It is expected that a full-scale prototype would comprise perhaps six or seven tapering WEC units rigidly connected side-by-side to form a platform with a roughly triangular plan view, and with sides up to 200 metres in length.
The platform would have a small waterplane area so that it remains steady in a seaway. The freeboard would be adjustable so that it was maintained approximately equal to the mean amplitude of the ambient waves. The deck would be flat, with possibly some relatively small, low-profile deck-houses. Storm waves could thus break over the deck and spill the major part of their energy so that WEC units would not be overloaded.
Initial deployment, to gain experience, would be in semi-sheltered areas where waves of excessive amplitude are not expected. Based on the predictions of the feasibility study, an output power of the order of 6 MW would be anticipated.
5.2 Synergies Since WEC platforms would be large, flush-decked and stable, they could accommodate wind turbines, ideally three per platform. These would substantially increase the power output from each platform, effectively sharing the overheads of the platform, the mooring system and the power-cable.
Figure 4. Platform Configuration |
5.3 Series Production
In series production, and possibly for the large-scale model and the prototype, it is expected that the WEC units would be constructed of concrete since increased weight is an advantage rather than a penalty.
In the production of twenty or a hundred platforms, considerable reduction in unit costs would be achieved, not only for the platforms, but also for the turbo-generators and control equipment. A flotilla of platforms could share a single power cable to the shore.
No new technologies would need to be developed for the construction or the deployment and operation of the platforms. | 6. OVERALL CONCLUSIONS |
With valued support from a DTI SMART award, a feasibility study for the OWEL WEC concept has been successfully completed.
Highly encouraging results were obtained from the tank-testing and mathematical modelling programmes conducted by QinetiQ Ltd. The final QinetiQ report predicted that a full-scale device would be capable of producing commercial quantities of power.
The
OWEL Company is now actively seeking funding to go on to a development phase which would include further mathematical modelling, leading to the design, construction and testing of a large-scale (c.20 metre long) model. It is expected that a development phase would confirm and refine the feasibility study predictions, leading to the construction of full-scale devices which would meet the criteria set out at the beginning of this paper in section 2.1
| 7. ACKNOWLEDGEMENTS |
