Kobold Technology Promotion and Transfer for Marine Current Exploitation in South East Asia. Papers and Presentations

OCCASION

This publication has been made available to the public on the occasion of the 50th anniversary of the United Nations Industrial Development Organisation.

DISCLAIMER

This document has been produced without formal United Nations editing. The designations employed and the presentation of the material in this document do not imply the expression of any opinion whatsoever on the part of the Secretariat of the United Nations Industrial Development Organization (UNIDO) concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries, or its economic system or degree of development. Designations such as “developed”, “industrialized” and “developing” are intended for statistical convenience and do not necessarily express a judgment about the stage reached by a particular country or area in the development process. Mention of firm names or commercial products does not constitute an endorsement by UNIDO.

FAIR USE POLICY

Any part of this publication may be quoted and referenced for educational and research purposes without additional permission from UNIDO. However, those who make use of quoting and referencing this publication are requested to follow the Fair Use Policy of giving due credit to UNIDO.

CONTACT

Please contact [email protected] for further information concerning UNIDO publications.

For more information about UNIDO, please visit us at www.unido.org

UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION Vienna International Centre, P.O. Box 300, 1400 Vienna, Austria Tel: (+43-1) 26026-0 · www.unido.org · [email protected] Kobold technology promotion and transfer

fpor lriAlali'llFilPB (C[Lrlli'rPBIFilIt (BZpalIpoiltatilpoiiii

ilail 5poUtlh Emote Atilt

Papers and presentations

UNIDO UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION Messina Conference 15-16 September 2005

Kobold technology promotion and transfer for marine current exploitation in South East Asia

Papers and presentations

UNl90

UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION Vienna, 2005 Copyright @ 2004 by the United Nations Industrial Development Organization

Designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the United Nations Industrial Development Organization (UNIDO) concerning the legal status of any country, territory, city or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. The opinions, figures and estimates set forth are the responsibility of the authors and should not necessarily be considered as reflecting the views or carrying the endorsement of UNIDO, The mention of firm names or cornrnercial products does not imply endorsement by UNIDQ.

This document has not been formally edited. Messina Conference 2005

CONTENTS

THE POLITECNICO DI MILANO ROLE IN THE RESEARCH "CURRENT FOR CURRENT" ALFREDO CIGADA, ALBERTO ZASSO

— 1 INTRODUCTION 7 2 — POLITECNICO DI MILANO IVIIVD TUNNEL 9 3 — ItVIND ACTION ON STRUCTURE 12 4 — THE KOBOLD VAIVE 15

THE ENERMAR SYSTEM ALBERTO MOROSO, HELENA ERIKSSON PONTE DI ARCHIMEDE

1 — INTRODUCTION 23 2 — THE DEVELOPMENT HISTORY - FROM AN IDEA TO 25 3 — THE KOBOLD PROTOYPE IN THE STRAIT OF MESSINA 30 4 - THE EIVVIRONMENT 35 5 - RESEARCH 35

ABSTRACT: A THEORETICAL AND COMPUTATIONAL MFTHODOLQGY TO STUDY VERTICAL-AXIS TURBINE HYDRODYNAMICS FRANCESCO SALVATORE, LUCA GRECO, GUIDO CALCAGNO, INSEAN

ABSTRACT 37 REFERENCE 39

EXPERIMENTAL AND NUMERICAL INVESTIGATION OF AN INNOVATIVE TECHNOLOGY FOR MARINE CURRENT EXPLOITATION: THE KOBOLD TURBINE G. CALCAGNO, F, SALVATORE, L. GRECO A. MOROSO, H. ERIKSSON INSEAN

ABSTRACT 4'I

SOME ASPECTS OF EDF MODELLING AND TESTING ACTIVITIES, WITHIN ITS MARINE CURRENT ENERGY RESEARCH AND DEVELOPMENT PROJECT ABONNEL C, ACHARD J, L, ARCHER A, BUVAT C, GUITTET L .. .

ABSTRACT 1 — INTRODUCTIOIV 43 2 - FRENCH "MARINE CURRENT ENERGY RESOURCES" AIVALIS YS 44 3 - VERTICAL-AXIS TURBINES EVALUATION: AN EDF-LEGI COLLABORATION 51 4 - CONCLUSION AND PERSPECTIVES 54 REFERENCES 55 GLOSSARY 55 Messina Conference 2005

FUEL CELL MICROCARS FOR SMALL SICILIAN ISLANDS V. ANTONUCCI, F,V. MATERA, L. ANDALORO, G. DISPENZA, M, FERRARO

ABSTRACT I - INTRODUCTION 57 2 — THE SICILIAN ISLANDS 58 3 - FUEL CELL VS, BATTERY 59 4 - MICROCAR DESIGN 60 5 - HYDROGEN fOR SMALL ISLANDS 62 6 - CONCLUSIONS 62 REFERENCES 63 AVTHORS 64

THE EUROPEAN COIVIMISSION'S ACTIVITIES IN SUPPORT OF RENEWABLE ENERGIES, IN PARTICULAR OCEAN ENERGY KOIVlNINOS DIAIVIANTARAS 65

Messina Conference 2005

The PoIitecnico di MiIano rote in the research "Current for current"

Alfredo Cigada, Alberto Zasso

1 - INTRODUCTION Due to the just mentioned facts, at this stage, literature mainly deals with gen- There is no doubt that the interest eral evaluations concerning the eco- towards renewable energies is going to nomic convenience to start operating grow in future years: the continuously "water mill" plants: the matter is to get growing trend in the petroleum cost the highest possible power levels, the will make this process faster and faster. best efficiency, to make comparisons Of course, according to each country against other energy sources, also to particular situation, this interest is find the best sites to build the new addressed towards different tech- plants. Of course also some sceptical niques, some of them related to sea approaches are encountered, but the current resources: under this point of authors remember similar facts when view Italy, surrounded by seas, shares "wind farms", for power generation with many other countries the aim to from the wind, first came out, then try exploiting this potential richness to spreading all over Europe. It's an impor- its full extent. tant occasion for both research and That's why Politecnico di Milano decid- technology growth and it can't be ed to pick up this new challenge and missed: staying behind now may mean started working in the topic of marine not to be able to fill the gap, with those current energy. The decision is rather believing now in the project, This is the recent, a nyway the state-of-the-a rt reason why the national and interna- research on this subject is still young, tional funding for research programs no commercial products exist today, so puts special emphasis on the renew- the moment has been considered as able energy systems, confirming the extremely favourable for a start. general interest on the topic: according Messina Conference 2005

to the preliminary documents, the EU rents and in some cases, like for Vll Framework Program has a special Indonesia, having a country broken in budget devoted to this topic and also a number of small islands. For these sit- the national research program has uations marine current energy extrac- recently opened a new call on the same tion looks very attractive, having a few theme. In addition it is remembered as if not at all competitors, as power has the Italian Association for Wind to be given to small communities (not Engineering is part in the organization so high rates) in places where cables for of the conference OWEMES 2006, next power transmission are hard to be year, in which Wind Turbines and Water brought. These countries have resolute Turbines are considered together in the will to get deep in this research, thanks work program. In 2004 the UK's to institutions like UNIDO, strongly Department of Trade and Industry has believing in the project, or the network announced an investment in wave and linking ASEAN and Europe. tidal power technologies of 75 million In this scenario also the co-operation euro. At the launch of the funding, UK between Politecnico and Ponte di Secretary of State for Trade and Archimede is worthy being mentioned, Patricia Hewitt said: "The UK's Industry, as both have shared common research wave and tidal flows are the greatest in interests in the past and still nowadays, Europe and l want to ensure we harness having a common factor, which is these immense natural resources to gen- water and water exp! oitation. erate power for the UK. Renewable ener- Remaining in the marine current har- — through wind, wave and other gy vesting, a patented device from Ponte sources - a vital in our fight plays part di Archimede, the Kobold turbine, has against climate change, and we are com- been taken by UNIDO as a prioritary mitted to further developing renewable project to be brought to Far East coun- energy to an increasing role in the play tries, and its improvement is the object UK's energy mix. This announcement of the present research carried out by reflects that vision and us firmly on puts Politecnico di Milano. the path to becoming the world leader in renewable energy" she added, The same This short summary traces the frame- statements ran be applied to Italy, or work in which Politecnico di Milano more generally speaking, to all Europe. research on marine current energy moves its first steps. The particular situation of Politecnico di Milano, having a lot of international As mentioned earlier, while literature links on a number of research projects, concerning a general economic evalu- and some important agreements with ation on this new energy source is institutions from all over the world, has rather rich, the number of papers con- offered the chance to get in close concerning both research and technologi- tact with some research groups from cal aspects is still poor, partly due to several countries, also in the far east, the young science, partly probably due which have expressed a strong interest to an attempt to preserve the intellec- towards marine current exploitation. tual property on the achieved results. These countries are China, the At this stage, Politecnico di Milano Philippines, Indonesia, all of them hav- research has just started, so this pres- ing special sites with strong tidal cur- entation is more a project plan, rather Messina Conference 2005

than a summary of results. Efficiency improvement by proper turbine design can take advantage of the well established experience This plan has been organized at two gained by the research levels: group operating since decades on the fluid elastic problems and now operating in the new large a) A first level concerning blade Politecnico di Milano optimization, aimed at considering the wind tunnel. Kobold prototype turbine, designed The use of a wind tunnel for a water and built by Ponte di Archimede, turbine optimization could be consid- already operating in the Messina ered rather odd at a first glance, how- Straits, Assuming this solution as a state ever the past experience of the of the art, any possible improvement research group has proven that once under the fluid dynamic point of view the main similitude scaling is account- is sought for, to make power genera- ed for, then the first stage improve- tion more efficient, ments are easier to be performed in air rather than in water, as instrumentation works in an easier and cleaner environ- b) A more general and wide ment. On the other hand the level, at which, according to the future hydraulic similitude is mat- availability of reasonable budget on approach rigorous no ter air or water is the flow up to air this research, the whole problem of flows speed at which compressibility marine current energy will be faced, effects are negligible. A similar relying on the different and high level was used with results in skills that Politecnico di Milano can approach good testing flood effects on offer within its Departments. At this bridge decks and on hull performances where stage the fluid dynamic optimization only the final stage was performed in water will be considered at a back-to-basic channel on an level: the vertical against the horizontal already roughly optimized model studied in wind tunnel, axis solution can be deeply re-considered, as well as the best kind of action It is worthy writing just a few lines on to be exerted on the vanes, drag or lift. this facility and the group working in it. Also many other aspects will be re-considered, like the power conversion and transmission, the layout of the whole 2 - POLITECNICO DI MILANO plan, as well as maintenance problems or the identification of suitable sites for WIND TUNNEL exploitation. Also full scale tests on the existing prototype could be planned to To the purpose of supporting by a get deeper into the problem. state-of-the-art facility the world-wide recognized excellence of Politecnico di Milano research in the field of Long- However, the adopted strategy is to Span Bridge Wind Engineering, as well focus on the first phase, limiting risks as general Aerodynamics, it was decid- and becoming familiar with the exam- ed to design and build a new large ined technology on the basis of a real Wind Tunnel, having a very wide appli- working plant. cation spectrum, very high flow quality Messina Conference 2005

standards and a number of testing facil- the vertical arrangement and flow cir- ities, The Wind Tunnel, being a refer- cuit are sketched in the vertical section ence all over Europe for Wind of Figure — 2 and Figure - 3, Engineering applications as well as for flow-structure interaction, is working at its full capability since September 2001 and in the four years of operations it has been fully booked for applications in both fields of Wind Engineering and el Aerospace applications. In addition to the already mentioned Wind Figure — 2: Wind Tunnel section Engineering projects, several Aerospace applications have been dealt with: Helicopters aerodynamics (Ag usta), Air intake aerodynamics, Flying model aeroelasticity, etc. . Figure - 1 shows an overview of the Wind Tunnel: it's a closed circuit facili- Figure — 3: Wind Tunnel section ty in vertical arrangement, having two test sections, a 4x4m high speed low The presence of two test sections of turbulence and a 14x4m low speed very different characteristics is peculiar boundary layer test section. to this facility, offering a very wide spectrum of flow conditions from very low turbulence and high speed in the contracted 4x4m section (I (0, 15% V =55 m/s) to the earth boundary tto litt) r layer simulation in the large wind engineering test section. Focusing on the boundary layer test section, its overall size of 36m length, 14m width and 4m height allows for very large scale wind Figure - 1: Politecnico di Milano Wind engineering simulations, as well as for Tunnel setting up scale models of very large structures including wide portions of The overall wind tunnel characteristics the surrounding landscape. The rele- are summarized in Table -1; vant height of the test section and its

Politeenico di Miltino 'Wind Tunnel very large total area (4m, 56m') allows Tunnel Oeerall Dimensions 50x15xl5 [m] for very low blockage effects even if 'Maximum Potrer (Fans onls) 1.5 [.tTIV] large topographic models are included. Mas Speed Turb. Int. Test Section The flow quality with smooth flow /s] I„% [m] [ shows a 1.5% along wind turbulence Bouudarr Larer 14x4 & 1.5 and 3% mean speed fluctuations in the Lo&r Turbulence 4x4 0, 15 ( measuring section. A very large 13m- diameter turntable lifted by air-film Table — 1: Overall wind tunnel charac- technology allows for fully automatic teristics rotation of very large and heavy model Messina Conference 2005

fitted over it (max toad 100.000 N). The The flow quality with smooth flow device is specifically suited for very shows a 1.5% along wind turbulence quick and easy change of wind expo- and 3% mean speed fluctuations in the sure of very long span bridge aeroelas- measuring section. A very large 13m- tic models, avoiding all the problems diameter turntable lifted by air-film related to the repetitive assembly and technology allows for fully automatic disassembly of those complex models rotation of very large and heavy model (Figure - 4). fitted over it (max load 100.000 N). The device is specifically suited for very quick and easy change of wind expo- sure of very long span bridge aeroe)as- tic models, avoiding all the problems related to the repetitive assembly and disassembly of those complex models (Figure -4). The wind tunnel has a float- Figure - 4: Wind Tunnel top view ing floor, allowing for a very clean modet set-up, leaving all the instru- The wind tunnel has a floating floor, mentation cable connections out of the allowing for a very clean model set-up, flow. The very long upwind chamber is leaving all the instrumentation cable designed in order to develop a stable connections out of the ftow. The very boundary layer and the flow conditions long upwind chamber is designed in are very stable even when parameters order to develop a stable boundary like temperature are considered, due to layer and the flow conditions are very the presence of a heat exchanger stable even when parameters like tem- tinked to the general control loop of perature are considered, due toThe the facility. The Wind Tunnel is operat- presence of two test sections of very ed through an array of 14 axial fans different characteristics is peculiar to organised in two stacks of seven 2x2m this facility, offering a very wide spec- independent cells. 14 independent trum of flow conditions from very low inverters drive the fans allowing for a turbulence and high speed in the con- continuous and independent control of tracted 4x4m section (I &0. 15% the rotation speed of each fan. This V =55m/s) to the earth boundary fully computer controlled facility can help in easily obtaining, joined to the layer simulation in the large wind engi- traditional spires 8 roughness tech- neering test section. Focusing on the niques, a very large range of wind pro- boundary layer test section, its overall files very different flow size of 36m length, 14m width and 4m simulating conditions and very different geometri- height allows for very large scale wind cal scales. engineering simulations, as well as for The wind tunnel process is fully con- setting up scale models of very large trolled a PEC and a computer net- structures including wide portions of by work (ABB control the surrounding landscape. The rele- system), monitoring through more than 100 transducers all vant height of the test section and its the most important flow parameters in very large total area (4m, 56m') allows terms of wind speeds, pressures, tem- for very low blockage effects even if peratures, humidity, vibrations of the large topographic models are induded. Messina Conference 2005

fans and of the structure, door opening allowed to set up measurement tech- etc, , and allowing for feedback control niques and correlated numerical simu- on the flow temperature and speed, lation models among the most The flow conditions were found to be innovative and advanced in the inter- very stable and a confirmation of this national contest, Figure - 5 shows the fact is the very low turbulence level in Messina Bridge section model being smooth flow. All the typical various set tested in the Wind Tunnel. The of spires have been developed in order Research Group is responsible for an to simulate the different wind profiles International Senchmark focused on and an original facility has been recent- the comparison of the methods pro- ly installed allotting for active turbu- posed by a worldwide network made lence control in the low frequency up by the most active and renowned range. research groups for the numerical simulation of long span bridge dynamic Concerning the low-turbulence high- response to turbulent wind. speed section, positioned in the lower arm of the circuit, the large dimensions (4x4m) and the quite high wind speed . (55 mls) enable to reach Reynolds numbers in the order of, The very low levels of turbulence reached in this section (0.15%), give the facility a very wide spectrum of possible applications. A number of transducers, instrumentation and data acquisition systems are available, allowing for all the typical boundary layer wind tunnel measuring applications in the wind-engineering field. Figure — 5: Messina-Sridge model - 3 WIND ACTION ON STRUC- The traditional up going research on TURE train dynamics, in the railway field, has The research activity in the wind-struc- been recently extended to the new ture interaction field has been first experimental data offered by the Wind focused on the dynamics of overhead Tunnel facility, allowing a deep under- power lines vibrating under the effect standing of the vehicle aerodynamics of vortex shedding, moving then to and mainly of the cross-wind effects. long span bridge aeroelasticity, still This subject is of crucial importance for now one of the core subjects of the very high speed trains: taking advan- research, The large expertise gained tage of the large sized low-speed sec- both in the experimental approach as tion, experimental campaigns on quite well as in the numerical one, resulted large scale models have been set-up in the availability of reliable simulation with a correct reproduction of different models for flow-structure interaction in boundary conditions related to the var- very different application fields. On that ious arrangements of the rail tracks subject, the new Wind Tunnel facility both in fixed and moving train config- Messina Conference 2005

urations. The experience gained in the Figure — 8 shows the scale model of an aerodynamic aspects, allowed, as a America's Cup boat fully instrumented consequence, to enhance the with 6 components force balance and Politecnico di Milano expertise in the 4 channels sails remote control. railway field resulting in a more com- Another research field is related to high prehensive and refined implementa- rise buildings and large flexible roofs. tion of the numerical models simulating the train dynamics, Figure - 6 shows an example of the instrumented and travelling train model in the Wind Tunnel. A research subject deeply integrated between base research and applications is about the dynamics of cylinders in the flow stream.

Figure - 8: America's cup yacht Figure - 6: Raikway vehicle ETR-480 Figure — 9 shows the scale model of a high rise building fixed to a 6 compo- Tensioned cables for very different nents balance and instrumented with applications, from overhead power 150 pressure for an lines to strands of suspended bridges taps, allowing understanding of the forces and pres- to submerged oil risers, are just some sure distribution due to the wind possible examples. Figure - 7 shows action. two tensioned cylinders in tandem a rra ngement in the Wind Tunnel Another relevant research subject is about sailing yachts.

Figure - 9: High rise building Figure - 7: Tensioned twin cylinders Messina Conference 2005

Figure — 12 and Figure — 13, on the other the Hurnber Bridge, measuring the hand, show examples of very large and response due to the turbulent wind. flexible roofs reproduced by aeroelastic scale models allowing to study the response of structures to turbulent wind.

Figure - 12: Flexible roof aeroelastic model

Figure — 10: Store-Baelt full scale monitoring

Figure — 13: Stadium roof aeroelastic model Figrue - 11: Humber Bridge full scale monitoring Experimentation needs a computational support, often needed to go from A CFD approach to fully understand the wind tunnel models to the full scale over mentioned subjects is now an prototype behaviour. Figrue — 10 and available tool, thanks to the increasing Figure -1'! show two relevant examples performances of computational of full scale campaigns managed by resources. Figure — 14 and Figure — 15 Politecnico di Milano, the first on the show results of simulations performed free standing Store-Baelt tower at the on bridges and trains, construction stage, measuring the vortex shedding excitation, the second on Messina Conference 2005

o Mathematical modelling o Experimental tests on the complete turbine

o Optimisation

The first phase, that is the evaluation - Figure 14: Multiple-box girder CFD of the existing situation, may be split simulation into different tasks:

o Setting up of a single vane model t'o be testedin the wind tunnel

o Setting up of a suitable dynamo- metric balance to measure the aerodynamic char acteristics o Wind tunnel measurements o Data analysis

The aim of the initial phase is to deal Figure — 15: ETR-480 train CFD simula- with the wind tunnel tests on a single tion vane. The key activities involve a detailed wind tunnel evaluation of the Experimental data are fundamental for turbine wing profile, measuring both the validation of the very promising the flow field in the surrounding of a CFD techniques, considering future vane and the global forces given by the developments of numerical methods stream. The allowance of a very large and computer science. wind tunnel section enabled to make experiments on a large scale model All these expertises are going to be put (2/3 real to model scale ratio) with very in the new project of the Kobold tur- low blockage level. bine, To achieve these measurements it has 4 - THE KOBOLD VANE been necessary to develop a dynamo- metric model of the turbine blade and run the necessary tests in the wind tun- Prior to starting a wide range activity, a nel, prior to come to the final elabora- research plan has to be outlined. As tion of the collected data and to the already stated, the leading idea is to overall evaluation of the existing equip- start with the existing Kobold turbine. ment efficiency. The overall process, which could lead In more details, the first phase consist- to the optimal adaptation of the tech- ed in setting up a vane model, for the nology to local conditions, could be wind tunnel testing operation. The planned in different phases. These vane model has been designed as light could be: as possible in order to limit the dynam- o Assessment of the present techno- ic effects on the recorded measure- logical development Nlessina Conference 2005

ments and at the same time it has been the fluid forces and the friction forces built very stiff, in order to reduce any exerted by the fluid on the body wind-structure dynamic interaction. exposed to the fluid stream. Therefore this operation has required Working under similitude conditions particular skills in the model construc- means that this ratio is the same for the tion (Figure — 16); model being tested in the wind t'unnel and the full-scale prototype. Advantage of large scale (close to full scale) wind tunnel tests is that the air cinematic viscosity is just 15 times lower than the water cinematic viscosity so that at Wind Tunnel air speed about 15 times larger than the corresponding full scale water flow Reynolds numbers very close to full scale can be achieved. This is the case of the studied Kobold turbine vane Wind Tunnel tests, where air Figrue - 16: building the vane model speed much larger than the full scale water current can be applied on a close the model has been built using carbon to unity scale ratio model. fibre, with a smooth surface paint- very With reference to the following sym- The is that of the ing. adopted shape bols and corresponding values: and existing Kobold turbine, the so called the fluid density, velocity and viscosity, "high lift 018". the vane chord, having indicated with the suffix "P" and "M" the quantities related to Prototype and Model, with reference to the standard I. S. values: p ArR= 1.83 10 P Water = 0. 0010 PAir = 1.23

0 C Water = 1.0 10 d P

the ratio between prototype and scale model Reynolds Number can be represented as follows: Figure - 17: the vane in the wind tun- pVB Re = nel jl

Special care has been devoted to simil- Rep P„,. . gI V~. BP itude scaling, as the real structure will „„,„, „„, operate under water, which implies 3f C 4A / 6f'arer .&i~ M. very high Reynolds numbers. Testing has been performed at the highest pos- P ] 4 87 rt'atd&' P sible Reynolds numbers (high speed, Re M Tj„,, B,~ large models). It is remembered that Having selected a scale ratio of the Reynolds number is the ratio between model in the order of 2 I 3 or in other Messina Conference 2005

words Bp l BM= 3 I 2, the Reynolds Number ratio prototype to model is in Rep ReM= V Vair. the order of / 22, 3 water l Being the typical current speed in the order of Vcurren t= 1 / 2 m/s, but with a relative speed under steady-state conditions around 2 times the current speed, then it has been assumed Vwater The next step has consisted in setting = 2 / 4 m/s, and being the tests per- up a special wind tunnel balance to get formed at a wind speed Vair - l 2 m/s the aerostatic vane coefficients. Forces the typical Re ratio will be in the order on the vane are directly measured, of Rep l ReM - 7.5 in the worst case. In together with the flow speed, in order other words the wind tunnel tests will to get the aerostatic coefficients for the be rather closely representative of the chosen vane profile. According to the full scale Reynolds Number, Special model scaling, internal or external load care will be taken to the surface rough- cells are usually adopted, The long last- ness problem. In fact it is well known ing experience gained by the research that a large roughness on the wing sur- group has allowed to fit a balance con- face can lead to earlier flow detach- cept, developed through years, to the ment at large angles of attack and special case of the "high lift 18 profile". consequently to lower maximum lift The balance consists of a couple of allowed by the lifting surface. On the external sensing elements, suitably other hand a larger roughness surface designed to measure even very small is typically representative of larger forces, having almost no hysteretic or Reynolds number conditions. The true friction effects, therefore producing vane surface, under operating condi- highly repeatable measurements. tions in the sea environment will very Part of the balance design is its calibra- quickly experience a growth in its, so tion, allowing to define measurement that it is believed that the very smooth uncertainty, also accounting for cross surface tests could be representative effects of the various force compo- just of laboratory conditions or brand nents, by defining a calibration matrix, new equipment. For these reasons instead of a simple calibration curve. both conditions were tested: smooth Tests have been repeated for three and rough surface Kobold turbine vane wind speeds, all the same it has been applying a very rough coating to the observed that this parameter has no wing profile, (see Figure — 18). effect on the static coefficient identification. Another aspect worthy being mentioned is the fact that the balance is part of a rotating device, in guise of a "spit", therefore allowing a complete 44 measurement on a 360' turn. The possibility of having data on the 360' rotation is an important fact, as the vane is Figure - 18 The smooth and the rough expected to work at whichever relative wing coating angle between the incoming flow and Messina Conference 2005

the vane axis: even if it is true that the functions of both the relative angle turbine steady-state speed is 3 times between the wind and the model and the marine current speed, then making are expected to exhibit some depend- the current-to-wing relative angle con- ence upon the Reynolds number too. It fined in a relatively low range, all the is hoped that no strong variation with same when the turbine starts its rota- Reynolds number is observed, and at tion this does not apply, and the 360' the same time the effects of a change data will offer a deeper insight in the in the surface roughness is investigat- start-up mechanism. ed. Concerning the pitch coefficient CM, its value varies upon the position Data analysis, at the present step, has in the profile section to which force consisted in getting the static coeffi- reduction is applied. In the present cients, The adopted references are approach it has been stated to consid- shown in Figure — 19. er a point at half chord, even if the usual choice in the aeronautical field is '/4 chord: simple calculations allow to switch from one position to the other. The wind speed has been measured by means of a pitot tube connected to a precision micro-manometer, directly giving the dynamic pressure term, while the rotation angle of the whole "spit" has been provided in two ways: through the stepped motor moving the rotating device and by means of a MEMS accelerometer. This one, under Figure - 19 Measure sign conventions quasi static conditions, measures the gravity acceleration component related The non-dimensional static coefficients to the device angular position accord- CD, CL and CM are defined on the basis ing to a cosine law. Finally, after an on- of the following formulae, considering site calibration verification, tests have that the aerodynamic forces are func- been carried out for the two roughness tions of the wind pressure and of a ref- levels of the wing surface. surface, conventionally erence wing Figrue - 20 to figure — 22 show the sur- assumed equal to the projected results obtained with smooth surface. face. ln the plots coefficients measured at coefficients are also The over defined different wind speeds (between 8 and 12 m/s) are shown together, as preliminary tests have proven that no differ- ence is appreciable due to this parameter variation in the mentioned field, Measurements have been repeated, even in different days, to ensure repeatability, and very low data dispersion has been observed. Data obtained in the surrounding of +180' rotation Messina Conference 2005

and -1800 rotation are also weil super- irnposed each other, confirming the 1.5 measurement set-up reliability. The plots are given both in the full range- 05 180'+180 as well as in the fundamental main working range of the wing -20 +20', In the former, the global behaviour of the wing is studied in its full range, also including the angle of 150 .II3I 50 0 50 IDI 150 200 attack regions encountered just occa- sionally, under turbine rotation start up Figure - 21: CL with smooth surface: or strong wake conditions. ln the -20 -180' +180' (up) -20' +20' (down) +20 range the effective wing operating conditions are studied in detail, 15 allowing a clear understanding of its ' / high lifting capabilities and of the corresponding low drag characteristics. As / from design, quite high lift values are / observed, in addition a number of . / / points evidencing sudden flow separation/reattachment are also observed 15 .10 8 0 5 10 15 20 out of the operating conditions; the ~ll stall at +12 and a singular point in CM at -10', again correlated to flow separation, are worthy being mentioned.

12 02 l'

00 O D5 , I 04 02

02 I SI 100 5D 0 50 100 150 XQ

150 100 53 0 50 IQQ 150 30 alN Figrue - 22: CM with smooth surface: Figure - 20: CD with smooth surface:— -180' +180' (up) -20' +20' (down) 180 +180' (up) -20' +20' (down) 02

015 03 0. 1

0.25

V 0. 15

4. 15

15 -10 8 0 5 10 15 20

tQ 15 -10 .5 0 5 10 15 20 Messina Conference 2005

The second series of tests has been per- Figure - 24: CL with rough surface: -180' formed on the wing with increased +1 80' (down left) -20' +20 (down) roughness. Results are given in Figure— 14 23 to Figure — 25, even if, for a better understanding of the roughness effects in the fundamental operating range of the wing, results from both test ses- sions have been then plotted together 06 47 in Figure - 26 to Figure — 28, 04 02

14 4 -' 022 .'', h l~P . ':M II 12 I, / -15 -10 4 0 5 19 15 20 9JJ

0 I

-' I

4 --Jg; ""4"--. 4-" {2 02 1

.Lp 150 100 -50 0 91 1{51 150 XQ 9 J'J 9 - with rough Figure 23: CD surface: 01 -180' +180' -20' +20' (up) (down) -I-- JI ', X ~ I, g 5I

93 O2

02 .ZS .I& .I{0 ~ 0 50 lQI 450 I 5 {0

0.15 Figure - 25: CM with rough surface:- 01 '"g "'! 180' +180' (up) -20' +20' (down)

-19 0 5 10 15 4 0 15 &JJ 01

15 / {{:I:/ / I Jh -"44' - ~, - I 05 j 1 15 16 W 0 5 10 15 I ~ I I

0 ~ J --4 ~ -- -- '- - ""I-. — '" ~ ~ -4l- " - ""-" Ol

100 50 0 50 160 {91 21m 20

Messina Conference 2005

Some facts are easily observed: as Mathematical modelling will be split expected, due to the roughness effects, into: encountered around a larger drag is o Model creation zero angle of attack and in the whole operating range, At the same time the o Parameter sensitivity analysis large roughness anticipates separation o Simulation of the complete t'urbine at lower angle of attack, resulting in a quite relevant reduction of the maxi- Experimental tests on the complete mum lifting capabilities of the wing. As turbine wilt consist of the following already mentioned the Reynolds num- tasks: ber range at which the tests were real- o Experimental set-up: complete tur- ized in the wind tunnel is somehow bine lower than that at full scale prototype conditions, so that it is expected that o Wind tunnel tests conservative in the tests are a bit the o Comparison between numerical way that the full scale wing, even with simulation and experimental meas- the same surface roughness could urements show a lower reduction in the maximum lifting capabilities and a separation a bit delayed to higher angle of Optimisation is going to be the final attack. part, relying on the results of the previous phases. A change is again observed when the wing offers its tail to the incoming flow, i.e. in the surrounding of -tSa', when the behaviour again is that of a wing profile. For high angles of incidence, when the wing works in stalling conditions, its behaviour is typical of a bluff- body section, being much less sensitive to the different Reynolds values, and therefore exhibiting aerostatic coeN- cients very close each other in the two test runs. The overall research plan has wide perspectives. Once the described preliminary steps have been run and the results have been achieved some further aspects are to be investigated.

To fully understand the real aim of the global work it is herewith given an out- line of the various phases or tasks within each of the above mentioned research phases: Messina Conference 2005

THE ENERMAR SYSTEM

Alberto Moroso, Helena Eriksson Ponte di Archimede

'I - INTRODUCTION axis turbines, compared to horizontal axis ones, are the designing and build- Marine currents represent a large ing simplicity and that the turbine, no renewable energy resource and have matter where the flow comes from, will the potential to give a significant con- always rotate in the same direction. tribution to fulfill the worldwide ener- - gy demand. The president of Ponte di The ENERMAR system with its core, — Archimede S, p. A. , Elio Matacena, came the patented Kobold turbine is suc- up with the idea of utilizing a vertical cessfully operating in the Strait of axis turbine to extract energy from the Messina since June 2001, marine current in the 80's. He had then In January 2005 it was drydocked for been inspired by how his ships, ro-ro maintenance, the generator and the ferries from Caronte S.p. A. shipping inverter have been changed to comply company, moved in the Strait of with the requirements for the Italian Messina, site famous since ancient electricity grid. In October 2005 the times for its very strong currents, by the Kobold prototype was connected to means of Voith Schneider vertical axis the grid. This is the first marine current propellers, very particular devices turbine in the world to be producing completely different from the normal electricity to a local electricity grid, see screw propellers -used for those ships Figure l. requiring a high maneuverability such as tugs and bidirectional ferries.

The main advantages of the vertical

23 Messina Conference 2005

current of 8 knots velocity (about 4 m/s), a normally strong current, the wind turbine will have to work in a

1 wind of about 136 km/h (38 m/s) which

' ~ KOSOLQ is the typical speed in a strong storm. '. i By this comparison it is evident that, at ' the same power output, the water plants are more affordable than the wind ones, smaller and implying minor building costs, and shorter payback Figure — 1: The Kobold turbine in the times. Strait of Messina As seen below, the produced electrical The theoretical power of any fluid flow power is the product of the global effi- (air, water, etc. ) is given by the follow- ciency, the turbine diameter, the blade ing formula: height (5=30 m' in case of Kobold turbine), the water density and the cur- P = 1/2 P 5 V' (P) rent speed (V). The actual efficiency— in which S is the projected area of the although it should more correct to call turbine, P is the fluid density and V is it power coefficient - of the Kobold tur- the current velocity. However, the max- bine is about 25%, This efficiency will imum extractable power is known as be increased further with an optimized the "Betz's Limit". This limit of 59.3% is mechanical and electrical system. For due to the fact that it is not possible to example the bearing of the turbine extract all the kinetic energy form the shaft was changed during the mainte- flow, which would stop the flow, nance and this increased the efficiency because a residual kinetic energy is of the system of about 3%. needed by the current in order to flow Furthermore it is worth a note that away from the turbine and leave space already the Kobold turbine has a effi- to other incoming flow. ciency comparable to wind ones, which have about forty years of development. In the theoretical power formula the velocity is present to the power of 3, if the velocity doubles, the power gets 8 times to double the power is bigger; KO enough to increase the current veloci- KO ty of the 25%, and so on. KO Another basic point of the formula is '-m P=OSxpx V xSxg represented by the fluid density: the Q ~ water density is about 850 times bigger Zj q =2s~/o than the air, thus having plants of the 1C0 same size, at the some velocity the 0 water plant will produce 850 times the 00 1,0 2,0 40 40 50 6j0 power of a wind plant; or also to have the same power of a water turbine in a, Nlessina Conference 2005

2 - THE DEVELOPMENT HISTORY FROM AN IDEA TO THE KOBOLD PROTOYPE IN THE STRAIT OF MESSINA

The very early tests of a current device commissioned by Ponte di Archimede S.p, A. were carried out in the hydrodynamic tunnel of the Voith, in Germany, in 1986, when several models were tested. All those prototypes derived from the Voith Schneider vertical axis marine propeller. ln the Voith Schneider propellers any blade rotates cyclically of a certain angle around its vertical axis; the law of this angle variation is related to an internal point: all the blades, during a revolution must have the normals of the chords directed towards this point. By changing this angle, by moving this point, it is possible to change the thrust direction in all the In this a 360. way Figure - 3: Voith Schneider hydrotur- ship propelled by a Voith Schneider is bine extremely maneuverable even with a single propeller and can sail without difficulty in any direction. The Voith turbine worked exactly in the same way, having the blades moving cyclically during the revolution. At the Voith hydraulic tunnel several models were tested having from 3 to 6 blades. The model which gave the best results was the 5 bladed one, showing the real possibility of Marine Current Energy, although there were some points which needed some improvement: the efficiency was not really high 3 (around 15%), the turbines in some conditions were not self starting and the Voith propellers (and therefore the turbines) are devices with complicated mechanisms and moving parts to allow the blades motion, thus heavy, delicate and expensive.

25 Messina Conference 2005

blades, thus without any mechanism to control the blade orientation, and having a high starting torque under any condition. Like any vertical axis mill, the Kobold turbine rotates in the same direction no matter the current direction, IK t The concept of a simple, cheap and reliable machine having characteristics of sturdiness and high efficiency was the i target of the study of Ponte di Archimede Co. which, in 1998 patented the Kobold Turbine.

A first mathematical model was then set up by Ponte di Archimede in order to evaluate the feasibility of this new concept. in this first prototype the blades of the turbine were modeled as flat plates and they worked in all the possible angles of attack, before and after the stall, thus having both lift and drag to generate the motive forces and the torque. The blade is not active in all its revolution but it has an idle fraction of revolution in which the generated forces are resistant and not motive, The first model of the Kobold turbine was designed by Ponte di Archimede and tested in the towing tank of Naval Engineering Dept, of University of Naples "Federico II" at the end of 1996 The blades were free to oscillate up to 90 degrees (with respect to the radial direction). Figure - 4: Voith Schneider marine propeller Two different models were tested in the towing tank, a first with 3 blades The Voith tests led Ponte di Archimede and a second with 5 blades. For both S.p. A. to start in 'l 995 the development of them the blade profile was a fiat of a new concept turbine which had to plate having a chord of 90 mrn and a be as simple as possible, without any height of 230 mm. The turbine diame- moving mechanism and, above all, self- ter was 800 mm and the two models starting in any condition. were tested at the velocities of 1.0, 1.5 The result of these studies was the con- and 2.0 m/s, cept of a new hydraulic turbine, the Kobold Turbine, having self-moving Messina Conference 2005

motion, thus changing the lift - drag ratio, improving the working fraction and reducing the idle fraction of revolution in the working point of the turbine.

Further improvements led the Kobold turbine to the actual configuration again protected by a new patent, A second numerical code developed at Dept. of Aeronautical Engineering of University of Naples "Federico II" was used to predict the turbine behaviour and output power, taking into account the interference between the blades- now evident for the higher rotational speed of the blades in respect of the Figure - 5: Kobold turbine model (5 first model —,the passive resistance of blades) tests in the towing tank of other turbine components (for instance

Dept. of Nava I Engineering of the arms) and other major aerodynam- University of Naples ic parameters.

The optimization of the turbine led to change some parameters in the working point of the turbine, so this time the torque was mainly generated by the lift of the blades and the idle fraction of the blade revolution was drastically reduced.

To validate the correctness of this new and more sophisticated mathematical model, the numerical activity was coupled with extensive experimental activities consisting in wind-tunnel tests of The towing tank tests fully confirmed a larger model of Kobold turbine. In the theoretical calculations of the first fact a new model was built and tested mathematical model of the turbine in the wind tunnel of Dept. of behavior. After this first stage of tests Aeronautical Engineering, University of the behavior and the working princi- Naples (see Figure - 6). ples of the turbine were more clear, so further theoretical studies on the blade motion brought the Kobold to a first patented optimization. The turbine performances were remarkably improved changing the angle of blade

27 Messina Conferenc:e 2005

tt)W t))If OIIMIII) PST )

Figure — 7: Wind Tunnel tests arrangement

A particular care was given to the pos- Figure - 6: Model of Kobold turbine in sibility to change the angle of blade the wind-tunnel of the Department of oscillation. The first Kobold turbine Aeronautical Engineering of University model tested had the blade oscillating of Naples (3 blades, up; 6 blades, down) like the ones tested in the towing tank. To optimize the angles and to avoid the influence of inertial forces on the blade oscillation, on the blade was positioned a counterweight in order to have the blades fully statically balanced, so to have only aerodynamical forces acting on them. The oscillation angle was controlled trough two adjustable wooden blocks used as limit stops for the blade itself (see Figure - 8).

)

) IAII ~ )IO I I II )0 1, This model was designed to work in the IIO 1 wind tunnel, thus working at completely different current (in this case wind} and rotational speeds and was built in such a way to change as many parameters as possible in the turbine configuration.

The model had a diameter of 2. 2 meters, blades height was of 0.8 meters Figure - 8: Details of the blade tip with and the blades chord was of 0.17 counterweight. Arrangements to opti- meters. It was tested with 2, 3, 4 and 6 mize blade pitch angle blades. The blade airfoil was a NACA 0018 standard profile and due to the In the following graphs a significant high number of possible parameters summary of the experiments is report- variation, hundreds of tests were per- ed in Graph — 1 it can also be seen the formed. effect of blades number on produced

28 Messina Conference 2005

rotor power. This gross graph clearly kobold .2 blared redr8 rrre shows why a 3-blade configuration was fired 82'

radaold Or renal 0'80 2 chosen for the real prototype (in fact P DARD kobold orrloel 0'861 180 lao bold 6 Prda ed irD' 90$ the maximum rotor gross power is the same of 4 blade arrangement, but with obvious less losses due to blade sus- taining arms and reduced construction costs),

kabobl - orlooal rkaok - ra\be

2 leadaa 0 40 80 120 180 2XI 2 rkadaa 260 280 66 ilko8 P iodk8 ~ badaa 1 liD 6 loaded Graph - 2: Kobold turbine optimization. Gross rotor power for different blade articulation settings (wind-tunnel tests)

gained: mainly it was eliminated the negative zone in the power curve, but also the total power output was D 40 20 120 1 BD 2D0 24D 28D re INror improved and the starting torque was increased, Graph - 1: Original Kobold turbine At high rotational velocities the behav- (blade articulation 0-90'). Gross rotor ior of the non improved model is power for different blade number con- almost the same of the improved one; figuration (wind-tunnel tests) the big influences of the way of the blade oscillation is evident in the turbine acceleration, in the way to arrive Optimization of blade oscillation angle was then performed to solve the prob- from the starting to the descending lem of negative power in the low rpm branch of the power curve, the usable branch for range. As shown in the first graph there the normal applications. is a small zone in the lower rpm region The in which the turbine suffers of negative turbine was tested several times, power output: to further accelerate and modifying its characteristics according pass this zone the turbine needs some to the numerical and experimental test external power. Once this transition has results. All the investigations led to the passed, again the turbine speeds spon- improvement both of the mathematical model taneously up to the top speed and of the turbine characteristics, deeper investigating the kinematical and In the Graph — 2 the gross rotor power the dynamic behavior for different blade oscillation angle set- of the tested device. ting is shown. Confining the blade oscillation angle into a very small sec- The final result of these theoretical tor (10 from the tangential direction) evaluations and model tests was the several advantages were Kobold prototype in the Strait of Messina, see Figure — 9. Messina Conference 2005

The Kobold turbine has been designed

I to satisfy, at the highest possible level, the environment safeguard and effi- ": p'a ciency needs, as well as the necessities of low construction and maintenance costs, The ENERMAR system has been designed so that minor causes cannot 'r result in disproportionately heavy damage. The design has taken into account the practicability of carrying out inspec- Figure - 9: The Kobold turbine in the tions of relevant components. Strait of Messina The characteristics of the Kobold tur- 3 -'THE KOBOLD PLANT )N THF bine are the following; STRAIT OF MESSINA - THE 0 direction of rotation independent ENERMAR SYSTEM of marine currenl direction

a very high starting torque, that ITALY makes the turbine able to start spon- iiPS'I taneously, also in loaded conditions, go~ R a p~'I without the necessity of any starting clevlces.

The airfoil used for the turbine blades 7 is a new concept unsymmetrical profile, so called HILIFT 'f8, designed for this purpose by the Department. of Aeronautical Engineering of the University numerical 7I Itaaarr of Naples using codes, taking into account both the maximization of the turbine performances and the risk of the cavitation which would quickly damage the blades.

~-~rpito, AA '~ I aollla laorrllAFa a It is the first time that was used an aaITatoao unsymmetrical profile for vertical axis Fiumoro titarolaa ~opttta l ~ turbine applications. The blades structure, mainly for the tra ~rII OII agaaflaa al P.to Ropglo ~ r 'rt. 7artapria hydrodynamic loads and for the :/ r ' / 'Ra. Cataana. weight, was studied using advanced -' atm arag. FEM programs and was realized in car- -Tromottl ji'i Re i- bon fiber and epoxy resin, dl c 'ria . I Figure - 10; The location of the Kobold prototype

30 Messina Conference 2005

'I'ttt Itt325525. 0 Pl,

Figure - 11: HILIFT 18 profile I 26~

155. 33'5 1 at 6 , I 3! /

I It 6 pj's/ )531 5 It' 5315 -!

ta 6563 55 25 6555 315 2513 15 )5 t)otaol5aa ~ Irta533566 Moat 5)15 1)aaaaall)56 35) Total Tiaaalaaoa - ooaaaa Loot!)aaaaaaa5aaaa 5 figure 14:- The ENERMAR system on the dock Figure — 12: Blade stresses The 3 -blades turbine rotor is mounted From the mechanical point of view, the under a round shaped buoy of 'IO m Kobold turbine has been designed fol- diameter. lowing simple and effective principles, The buoy was built in steel according so as to need for its whole useful life to the Italia Shipping Register (RINA) very limited maintenance interven- regulations for the steel ships and cer- tions. The design has taken into tified by RINA. account the practicability of carrying The main characteristics of the floating out inspections of relevant compo- platform are the following: nents. The main turbine dimensions are the Diameter 10, 0 m following: Depth 2.5 m Design Draft 1.4 m Steel weight 25, 0 t Diameter 6 meters Displacement 35.0 t Blade Span 5 nleters A)I etacentric height m Chard 0.4 meters 5.0 N' of Blades This last parameter indicates the stability characteristics of the floating platform: the bigger is the metacentric height, the higher is the platform stability. If compared to a standard ship with the same disptacernent, the platform stability is about 6 times bigger. In naval architecture the metacentric height is indicated with the term (r-a) in which "r" is related to the geo- metncal parameters of the floating Figure - 13: The ENERMAR system platform (immersed volume and moment of inertia around the inclining Messina Conference 2005

axis of the waterplane), while the term "a" is related to the floating platform mass properties and indicates the ver- REACTION tical position of the center of gravity. To have a big stability, thus, it is important to have the term (r-a) higher as RUSI' possible, and it is normally achieved keeping the center of gravity lower as possible. Another main point for the stability — moment in work- characteristics of the floating platform Figure 15: Trimming ing conditions is represented by the displacement 4 (equal to the platform weight), in fact is very closed to Ganzirri, in the Strait the stability moment is given by the fol- of Messina, by the Sicilian coast, distant lowing formula in which 9 is the incli- from the shore about 150 - 200 m. nation angle: The depth goes from 15 to 35 m and the maximum current speed is around (r-a) sin = Moment of Stability 5 (} 2.0 rn/s although there are places, in the Strait of Messina, where the current Since the turbine, when working, pro- speed can be more than 3.0 m/s. duces a thrust having approximately the current direction, this thrust (which The moorings, due to the weight of the - is around 10 15 t) generates with the lines, work like non linear springs but mooring reaction a moment inclining exponential, following the law of the the whole plant. catenary. This implies that if the dis- obvi- This trimming moment depends placement of any line it is a little more ously on the thrust but also on the arm, than the others, the force of this line mid-span i.e. the distance between the will be very much higher than the oth- of the blades and the mooring point on ers, thus the mooring lines work only the platform, one per time, and this is truer the effi- For reasons of both global plant tighter is the lines, i.e. the higher thrust ciency and safety on board, it is impor- of the turbine is. tant that this trimming angle is as small in normal as possible. Actually, working The mechanical energy produced by is conditions, the trimming angle the turbine is turned into electrical by around 5 degrees. means of a synchronous brushless 380 V three-phase, four poles electric gen- The platform is moored to the seabed erator. Since the turbine rotates very means four mooring lines com- by of stow (18 - 20 r. p. m. ) while the genera- posed each of a chain {27 m) at the sea tor, to have an output at 380 V and 50 bottom and of a textile rope going up hz, needs to rotate at 1500 r, p. m, , the to the platform. The anchoring devices turbine is connected to the electric are 4 mooring blocks made of concrete generator through an epicycloidal having the weight of 35 t each. gearbox with ratio 90:1 increasing thus the rotational speed at the generator The site where the plant is positioned shaft.

32 Messina Conference 2005

The main losses that are present in the system are mainly of three typologies: mechanic, electric and hydrodynamic. There are mechanical losses in the bearings and in the gearbox, there are yL electrical losses (the so called losses in the ncupper") and ventilation losses (for cooling) in the generator and there are hydrodynamic losses in the bracing of the turbine. These last ones are "physi- ological", they can hardly be reduced and this can be done only changing Figure — 16: Machinery room some design parameters. Generator and gearbox

The global efficiency of the system, also and more correctly called power coefficient, is defined as the ratio between pm + the produced electrical power and the +

theoretical power available in the cur- +

+ rent relative to the intercepted area: +

Pe lectrica I + .SpV S +

2 06 s oo 6.66 6. 00 Tampa [h) where S = Diameter' Blade Height (5=30

m2 in case of Kobold turbine), is — P Figure l 7: Current velocity in the Strait water density and V is the current of Messina speed. The measured global efficiency was (before 2005) around 23%, which There are some "added" hydrodynamic is comparable to the long time well losses which take origin from the blade wind turbines developed and so this braces when the whole plant is inclined first results can be considered excellent during the normal working. For this even because on-going improvements inclination the brace fairings don't work in the mechanical transmission system any more at zero angle of attack but will certainly rise the global efficiency with an angle equal to the inclination soon. very angle, so they generate a lift and thus an added drag which represent a pas- The power produced was calculated sive resistance for the whole system, to measuring the current in every phase be kept as low as possible. and the tension between the phases, therefore the power measurements Although the plant design power is were net and didn't take into account about 80 kW, the maximum power out- the several losses in the various put was around 25 kW due to the devices. installation site, which is not the best,

33 lNessina Conference 2005

—— V='&40mts — —V=1600ss for this purpose, in the Strait of ~V=120mis ~V=!37mts Sr ~V=1000Vs a Messina, site famous among the sailor- 2 4 I' ~k men since ancient times for its strong 20 and dangerous currents. In fact the cur- K~ 16 rents are generated by the phase oppo- 0 sition of the tides in the Ionian and in 12 the Tyrrhenian seas, therefore every 6— 40 7 hours there is an inversion in the current direction. The two currents are 9 19 known as discendente (descending, Shaft Angular Speed (rprn) - Tyrrhenian Ionian direction) and mon- Figure — 18: Kobold power output tante (rising, Ionian — Tyrrhenian direction). plant has just been permanently con- Although in some points of the Strait nected to the italian electric grid, This the current can reach more than 3 m/s, is the first marine current energy plant in the place where the plant is moored to be permanently connected to a the current velocity is hardly more than national grid, 2 m/s and the plant is sheltered from the descending, thus it is active only Since the turbine rotates at any speed, during the phases of the rising current. depending only on the current velocity, the electricity is produced at any fre- Up until January 2005 the electricity quency and tension. This gives no produced by the turbine was used on problems vvhen the electric power is board only for experimental purposes, used on board for experimental pur- by turning on 20 floodlights each of poses, but when giving electricity to them absorbing 1 kW of power (total the national grid, the requirements are produced power 20 kW) with a current very strict in terms of frequency, ten- speed of about 1.8 m/s and driving the sion and phase. electro-pump (25 kM/) with a current speed of about 2.0 m/s. In this way, The energy is produced by the system except the pump, the electrical load at any angular velocity and, since there was only resistive, purely homic, so it is no any blade-pitch control, the tur- was possible to calculate the power bine rotates at any velocity, depending multiplying the current in the phases only to the current speed. The only way by the tension between the phases. to control the r. p. m. is to change the load of the turbine by keeping it at the In a later stage a computerized system maximum power output for that cur- was installed on board to continuous rent speed. Since the electrical current monitor several parameters of the sys- produced is directly dependant, in tem, a torque meter was added to the terms of tension and frequency, on the generator shaft to calculate the power angular speed of the generator (and before the generator, thus non includ- thus of the whole system), it is neces- ing the tosses in the generator in the sary to regulate these parameters power output of the system. before sending the energy into the national g rid. As reported in the introduction, the To meet the grid requirements a static

34 Messina Conference 2005

rectifier-inverter was installed on board 5 - RESEARCH in order to have, no matter the turbine speed, always the same electrical out- With the experience from a full scaled put in terms of tension and frequency, system in the water for 4 years of time and always in phase with the national Ponte di Archimede S.p. A has all the g i'Icl, necessary information to develop an optimized mechanical system and is V = VARIARLE f = VAR IABI E further on developing a design tool together with INSEAN in Rome and Politecnico di Mila no.

TO THE GRID EXTERNAL — SETTINGS V=88OV Since the transfer air water is not so f = 80 Hz immediate, the contribution of INSEAN, with its experience in naval architecture, both numerical and experimental - Figure 19: Main scheme of the electric (INSEAN has one of the biggest towing part tanks in the world), and of Politecnico di Milano, with its experience in aero- 4 -THE ENVIRONMENT dynamics and wind tunnel tests, will be precious to fully understand the differ-

The environmenta I impact of the ences in the turbine behavior in air and Kobold turbine has been evaluated par- water. This comparison will be very ticularly from the point of view of the important to decide which studies can compatibility with the sea, flora and be carried out in the wind tunnel - wind fauna. The environmental impact and tunnel tests are cheaper and easier to compatibility study, carried out by the manage in case of multiple configura- University of Messina (ITALY), has tion models — and which ones have to reached the following conclusions: be necessarily done in the towing tank or in the hydrodynamic tunnels. It is o the environmental impact is negli important to note that the tests carried gible. out in the wind tunnel are in a completely different — o the Kobold units are compatible range of speeds just make with the Italian rules for theinstal- to a comparison the Kobold at sea rotates lation and removal of sea struc- at less than 20 r. p. m. , the tures. model tested in the wind tunnel at University of Napoli had an angular speed of more than 300 r. p. m. —and this The visual impact of the plant is very can cause significant differences in low and, case of plant removal, the between the model tests and the full works for the site reclamation are of a scale realizations because of the influ- small amount: there are no permanent ence of other phenomenon. structures on the seabed and the bigger components, the mooring blocks, At present both INSEAN and are easily removable using a normal Politecnico di Milano are working on crane. the project, with Politecnico di Milano having already tested the complete

35 Messina Conference 200S

polars of the airfoil used for the blades In the towing tank tests a couple of of the Kobold, and INSEAN ready to test Pitot pipes will be positioned in differ- a Kobold model in different configura- ent places to investigate the influence tions using its towing tank facilities, of the distance from turbine on the measured velocity, since it is exactly

e known the flow velocity which is the carriage velocity.

Furthermore, using the most advanced techniques in computational fluody- namics (CFD), a new sophisticated mathematical model is near to be ready in order to have a more powerful tool to estimate different design conditions without carrying out expensive model testing. This will be very helpful for future projects where the turbine will have to be optimized for specific site conditions. The computer model and the optimized mechanical and electrical are to be finished in 2005. Figure — 20: system Kobold model P at IN5EAN

Another important point that will be tested in the towing tank is the influence of the turbine itself on the undisturbed flow: it was noted, during the past tests carried out on board, that the current velocity measurements are influenced by the turbine. If we measure the current velocity too far away from the turbine we measure another current which could be significantly different from the one interesting the turbine, If we measure the velocity too near, there is the turbine influence, thus the measured current is slower than the real. It is important to investigate this phenomenon otherwise the efficiency calculations could be not effectively real. Messina Conference 2005

Abstract: A theoretical and computational methodology to study vertical-axis turbine hydrodynamics

Francesco Salvatore, Luca Greco, Guido Calcagno,

INSEAN — The Italian Ship Model Basin (Italy3

ABSTRACT merged below a floating platform moored to the sea bed. The blades The current energy demand and its dra- have a rectangular, high-span planform matic rate of growth related to the and are free to oscillate around a span- impact of developing countries, clearly wise axis (variab(e-pitch mode). indicate the necessity to consider A research activity aiming to develop renewable energy as a primary mean to computational tools to predict the alleviate the needs of a large number hydrodynamic performance of a of communities all over the world. 'Kobold' turbine is under progress at Among renewable energies, marine INSEAN, in the framework of a collabo- currents represent a relatively new and ration with 9onte di Archimede S.p. A, , almost unexploited source with a the Italian company that develops this worldwide diffusion of potentially high- particular hydroturbine concept. ly-productive sites. Recent research on both horizontal-axis and vertical-axis In order to provide a realistic model of hydroturbines have demonstrated the the turbine flow under operating con- feasibility and advantages of power ditions, a very general hydrodynamic generation plants based on those can- model is proposed. This is a distinguish- c epts. ing feature of the present work, in that classical hydrodynamic tools used to Scope of the present work is to address estimate the performance of vertical- the hydrodynamics of a power genera- axis hydroturbines are typically limited tion plant based on the Kobold turbine to simplified models as blade-element concept. This term denotes a vertical- theory and isolated vortex methods axis multi-bladed rotor that is fully sub- (see, e. g. , Pawsey, 2002). Two major Messina Conference 2005

aspects affecting the turbine hydrody- multi-bladed turbine studies, the trail- namic performance are deeply investi- ing wake model allows to account for gated here. One aspect is related to trailing-vorticity/blade-surface interac- predicting the blade kinematics in the tions. Viscous-flow contributions to tur- variable-pitch operating mode, The bine torque and power are estimated second aspect addresses the interac- through equivalent flat-plate approach tions among the turbine blades and the to evaluate the friction drag on the sur- effect on the performance of blade- face of the blades. generated wakes approaching and impinging on the blade surfaces. ln order to assess the proposed methodology, the present research also In order to focus on these two aspects, includes an experimental activity aim- two theoretical models are separately ing to provide benchmark validation developed and then integrated, data. Specifically, model hydroturbine First, a blade dynamics model is used tests at the INSEAN towing tank facility to determine the blade pitch angle as are planned to characterize the turbine a result of the hycfrodynamic forces act- working conditions (rotation speed, ing on the blades during a turbine rev- torque, power) over a wide range of olution. A simplified quasi-steady variation of the simulated current hydrodynamics approximation is used speed. A detailed review on this exper- in which lift, drag and moment of a iso- imental activity is presented by lated blade in uniform translation at Calcagno et al. (2006). given angle oF attack are prescribed. Blade pitch angle variations along a tur- Three-b4ded Kobold turbine: bine revolution are then obtained by Three-dimensional view of computa- studying the response of an equivalent tional grid on blade surfaces (in red) mass-spring-damper model. and trailing wakes (in cyan). In parallel to this, a general unsteady hydrodynamic model of a multi-bladed rotor assembly with prescribed blade pitch is developed. By combining the blade dynamics model and the turbine unsteady hydrodynamic model, an integrated performance prediction tool for variable-pitch operating turbines can be derived, Both quasi-steady and full three-dimensional unsteady hydrodynamic models are based on a boundary element for- mulation that is valid for potential flows In the final paper, numerical predic- (see, e, g. , Greco et al. , 2004, addressing tions of the performance of a real- marine propeller hydrodynamics), shape hydroturbine will be presented Vorticity shed by the blades and asso- and compared with experimental data ciated to hydrodynamic force genera- from towing tank tests. tion is described through a suitable trailing wake model. In the case of Messina Conference 2005

V LoCAI

~ .rr

Description of the turbine kinematics and hydrodynamic forces in a horizontal plane.

REFERENCES

Calcagno, G, , Satvatore, F., Greco, L, , Moroso, A. , Eriksson, H. (2006), 'Experimental and Numerical investigation of an innovative Technology for Marine Current Exploitation: the Kobold Turbine, 'submitted to the 2006 Conference of the International Society of Offshore and Polar Engineers (1SOPE).

Greco, L, , Salvatore, F, and Di Felice, F. (2004), 'Validation of a Quasi-potential Flow Model for the Analysis of Marine Propellers Wake, ' proceedings of the Twenty-fifth ONR Symposium on Naval Hydrodynamics, St. John' s, Newfoundland (Canada), August, 2004.

Pawsey, N. C, K, (2002), Development and Evaluation of Passive Variable-Pitch ' Vertical-Axis Wind Turbines, PhD Thesis, University of New South Wales, Australia.

39 Messina Conference 2005

Experimental and Numerical investigation of an Innovative Technology for Marine Current Exploitation: the Kobold Turbine

G. Calcagno, F. Salvatore, L. Greco A. Moroso, H. Eriksson

lNSEAN — italian Ship Model Basin

ABSTRACT building simplicity, sense of rotation independent from the current direc- IVlarine currents represent large renew- tion, a very high starting torque and a able energy resources anct have the high efficiency. A prototype has been potential to give a significant contribu- realized and it is now successfully oper- tion to fulfill the worldwide energy ating in the Messina Strait in Italy since demand. 2001, producing, for the first time in the The possibility to extract power from whole world, electricity from marine both air and water currents is well currents delivered to a local electricity known to the human beings since g I'Id. thousands of years, and modern technology makes the exploitation of this In this paper we describe the activity in type of renewable energy more and the framework of a research project, more reliable, safe and cost-effective. Aims of the project are to achieve a An innovative technology for marine better understanding of hydrodynam- current exploitation based on the ics aspects involved in the Kobold tur- Kobold turbine concept is addressed bine operation and to develop here. Basically the Kobold Turbine is a prediction tools to assist design of vertical-axis multi-bladed rotor that is high-performance Kobold turbines. The fully submerged under a floating primary role of hydrodynamics studies moored buoy and which is subject to stems from the fact that the successful an incoming flow originated by a tidal design of a Kobold power generation sea current. This kind of turbine has plant is necessarily related to maximiz- several advantages compared to the ing the energy transfer from water to horizontal axis ones: designing and the shaft through the turbine blades. In Messina Conference 2005

the present work, both experimental one. techniques and theoretical/computa- In the final paper, the experimental tional models have been used to methodologies as well as the theoreti- address such aspects, cal models used in the project will be described. Turbine performance meas- In order to determine strategies to opti- urements from towing tank tests will be mise the turbine's design we have test- analysed. Next, numerical predictions ed different models of Kobold turbine by the proposed computational tool at the INSEAN towing tank, simulating will be presented and validated the marine current with the turbine through comparisons with experimen- towed by the carriage. Measuring tal data, torque and angular velocity for each current speed and for different number of blades, we have defined the overall hydrodynamic performances in terms of the delivered power on the shaft. These experimental results on the model have been compared with data available for the full scale prototype recorded along its operational life on the sea-site.

From a theoretical and computational perspective, a fully three-dimensional unsteady hydrodynamic model has been developed based on a Boundary Element Method. This methodology allows to include in the performance Sketch of a three-bladed Kobold tur- analysis the unsteady effects of the vor- bine tical wake shed by each turbine blade V and to understand the influence of the hlurrr b Effective force: —F cohP blade-wake interaction on the torque F,rr generated by the turbine. Torque: Q = F, x R The theoretical model has been imple- Power: P= Q yc Q mented into a hydrodynamic performance prediction software that has a low computational cost and thus repre- F = hydrod ynnrnic force ou triode sents a practical design-oriented tool. « = elude;rnfler of nttnck

As a matter of fact one of the major J3 = qtl' —« problems with such a kind of turbine is that it has to be site-tailored according to the different marine current conditions which, albeit largely available worldwide in coastal regions, can greatly differ both in intensity and in time duration from one site to another IIIiessina Conference 2005

Some Aspects of EDF Modelling and Testing Activities, within its Marine Current Energy Research and Development Project

ABONNEL C, ACHARD J.L, ARCHER A, BUVAT C, GUITTET L,

LENES A, MAITRE T, MANIATI M, PEYRARD C, , RENAUD T, VIOLEAU D

ABSTRACT for several decades now, as the La Rance power plant - 240 MW installed, This paper discusses some aspects of was launched in 1966, near Saint Malo the modelling and testing activities in Brittany. While this plant goes on deployed by the Research and producing electricity today, EDF got Development Division of EDF (EDF engaged at the end of the 90's in new R&D) and one of its French academic research and activities on complemen- partner, the LEGI laboratory, to con- tary means of production, aiming at tribute to the development of a novel diversifying its sources of clean renew- technique for power generation using able energy. Among such sources: the the kinetic energy of tidal streams, still ocean or marine ones. Indeed, France is called Marine Current Energy (MCE). well endowed with both wave energy Through numerical modelling, experi- and marine current energy (MCE)' mental measurements at sea and tank according to the EC sponsored study testing, EDF improves its knowledge of "Marine Currents Energy Extraction: the French potential tidal sites and also Resource Assessment" (CENEX, 1996), its capacity to evaluate the various France has got 20% of the global technologies and concepts some of European marine current resource— which could become industrial tornor- estimated around 48 TWh for an row. installed capacity of 12500 MW, a little more than the 10th of the world 1 - INTRODUCTION resource estimated around 450 TWh, following certain pre-defined charac- Electricite De France, the major French teristics to make them suitable for ener- utility, has been producing tidal power gy exploitation, Between wave and Messina Conference 2005

marine current sources„ the choice was different scales, from the "large" picture made to focus on MCE, as it presents of the whole Close Atlantic Channel several interests: potential contributor and North Sea area to the converter to the base load, predictable, dense vicinity, going through the farm area, and regular, it is also less exposed to each time dealing with both numerical severe storms than other marine and experimental data. Similarly, to renewables. evaluate the developing technologies, we wIII present the collaboration with Thus, EDF launched its R8D MCE proj- the LEGI, based on both experimental " " ect Hydroliennes en mer in January and numerical approaches. 2004, aiming at: (i) evaluating the maturity of technologies being devel- 2 - FRENCH "MARINE CURRENT oped within this emerging tidal market, ENERGY RESOURCES" ANAL- (ii) identifying and characterising the ISVS tidal sites in the French territorial Although the IVICE waters, in order to prepare the deosion resource is potentially enormous, in of the EDF Group to invest (or not) in a most sea areas the velocities, and first industrial demonstration site in the hence the energy densities, are too low French territorial waters by some years. for economic exploitation, Therefore One have noticed that EDF, as an may the resource of interest has to be veri- industrial utility willing to become an fied and refined as there are only few end-user of IVICE, focuses its R&D activ- places that should be usefully exploit- ities on improving both its knowledge ed: places where the seabed topogra- of the potential French tidal sites and phy and the effect on concentrating its capacity to evaluate the various the flow caused by coastlines, such as technologies and concepts being in straits and around headlands, cause developed. Thus, while other entities, fast currents to occur regularly. as the ones presenting in this first CA- GE workshop, deal with the develop- 2. 'i - The "large" picture: modelling ment of devices themselves, EDF and measuring tidal currents chooses a complementary approach, by itself or in partnership, to "pull" this 2. 1.1 - Numerical modelling with the emerging market of ocean energies. TELENiAC system To do so, several tools are used accord- In order to make a first evaluation of ing to the issue. Some examples are the marine current kinetic energy on provided in this paper: numerical mod- the French coasts, to compare with the elling and measurements at sea to deal results of the European study (CENEX, with the resource, numerical modelling 1996), a pre-existing numerical 2D- and tank testing with the LEGI hydrodynamic model was used at EDF, Laboratory to evaluate various con- based on the 2-D shallow-water equa- cepts. One should notice that the tions (or Saint-Venant equations) which KOBOLD turbine of Ponte di are known to be suitable for modelling Archimede is not considered at this tides in coastal areas, The model, con- stage, but it could be in the future. structed with the Telemac-2d code, developed in a finite element formal- Thus, in this paper we will present three ism at EDF R&D for tidal and river appli- Messina Conference 2005

cations (Wervouet, 2003), lies from the North Sea to the Spanish coasts including the Channel and a part of the ««naac&W

Atlantic Ocean (Figure 1). The mesh is &&.&& ~-eo composecf of 15,350 nodes and 29, 229 &-~o & -~.o triangle cells. The conditions U -&oo type input &-~o at liquid boundaries are tide signals o -&&0.0 -70&& which are generated in the Atlantic D ~a D Ocean, move over the European conti- 0 -&00« + «&0« 'I -&OOOO nental plateau and penetrate into the -~.o ~ -~0 Channel.

French coasts. Pretty good results were obtained in terms of water levels and velocities, provided a sufficient distance from the simplified coastline is respected. Such results were coherent with the European study.

To make a realistic assessment of the energetic potential of the coastal areas, tides of a whole year were simulated. It was assumed that one year was sufficient to mask the inter-annual tide coefficient variability, and therefore give representative results. The output variables were velocity and kinetic instant power (Figure 2), Annual extractible energy was obtained by integrating the instant power over the year, Theoretical extractible energy was considered. It is given by the Betz law which corresponds to an ideal turbine in an infinite medium:

With;

Figure 1 - Bathymetry (right) and mesh of the North Sea, Channel E g extractible kinetic energy and Atlantic Ocean "Large TELEMAC (j/m') Model" (up) P = sea water density (kg/m') V = velocity (m/s) The model was validated with tidal t = time (s) data from several harbours along the Messina Conference 2005

This gross energy will then have to be straints. From this list, the more inter- converted into effective extractible esting sites were selected as potential energy by considering the selected tur- project sites where a turbine prototype bine technology and the correspon- could be installed. Thus, the GIS system ding efficiency. Other effects as turbine constitutes an achieving tool that can wake in farms will also have to be taken be used and updated with the detailed into account, as detailed in 5 2.2. studies engaged from there: experimental measurements at sea {see 5 2. 1.2), smaller scale numerical models to go further with complex currents cir- culations and therefore allow "tidal farm" 2, v 1 O111 design {see 5 2), (OO/O) The boundary conditions of such local models are supplied by the "large" 11 OOOO telemac-2d model results.

o 0. 6 712 - Experimental measurements at 0 I 1 2. @ a. . ~ ~ sea

-OOOOOO -OOOOOO -oooOOO -OOoooo -OOOOOo -OOOOOO I OOOOOO OOO'OOO

Figure 2 — instantaneous velocity along Numerical models previously present- the French coast, obtained with the ed reproduce only tidal currents, Telemac-2d large model. depth-averaged and without any waves at the free-surface. Real marine Once the energetic 'crude' information currents are actually complex flows on is available, the second step towards a which free-surface effects have a signif- site identification consisted in crossing icant influence. The vertical shear for a ail these calculation results with envi- current turbine is typically larger than ronmental and regulation data, A for a wind turbine, inducing cyclical Geographic Information System (GIS), loading. The addition of waves further based on the ARCVlEW software, was complicates the flow, with its influence used. Once the data were georefer- being more than just an adding-turbu- enced, the GIS system permitted not lence effect. only to superimpose the different data layers, but also to make surface and As few data actually exist for the water mean calculations. The cross-analysis of column itself in such zones of strong all the defined constraints, such as currents, we decided to have experi- energetic potential, distance from the mental measurements at sea for sever- shore, bathymetry, military or civil reg- al reasons: ulated zones, marine activities etc, leads to the identification of roughly to confirm the local validity of our ten potential zones along the French numerical results obtained with our coasts. large model in the most interesting zones: this is discussed straightfor- These zones were sorted according to ward their energetic potential, their technical feasibility and their environmental con-

46 Messina Conference 2005

o to gather data to help understand Point n' Tr e of decice(1 !rfeurured nrnmeter1 fmmertion de tb waves ARK'AC tVnee oud current 0 n1 how the can influence the ADP dOO l'Hz Current IO nl velocity in the water column. First ADP f00 bHz C urf cut 0 \'I'I Aqu tdopp profiler 6(10 Current zl nt elements discussed in 5 2. 3.2 l'Hz Dntntc«ll 1corebuor i't'uue 1ur fne« tVuue nud current 00 n1 o indirectly to "open" the discus Table — 1: Position and type of rneasure- sion and consultation with other ment devices used during the January sea-users and authorities about the campaign in Normandy. Except for the development of MCE, as acceptabil- Datawell buoy, all the other devices are ity will be one the key-factors of a designed by the Nortek company'. successful deployment.

The first campaign was achieved in January 2005, in the so-called uRaz de Barf leur", in the North-East part of the Basse-Normandie department (Figure- 3a). ~"Wi ~~ '

r -." 0--1 '0 , f , "4

1 ' J X p 3 ~ o Q mb&t ru e*arnue

et ~ .1~ - -. 5 Figure 3 b: Example of device installa- e tion/removal on the sea bottom. en r r 1 All these devices are based on the Doppler principle'. Programmed at Figure — 3 a: Location of the devices the beginning, they are fully autonomous during the Normandy measurements in for measurements and of January 2005. The modification of the storage data in an internal datalogger. ADP and point 1 illustrates the "consultation AWAC measured process" with local fishermen. velocities every meter in the water column, averaged 1 or 1.5 minutes The measurement devices remained every 10 minutes; the Aquadopp in point 4 measured veloci- averagely 15 days (a whole tidal cycle) ties every 2 m, averaged 1s every in the water, between 7th January and 3s. With the measurements val- 5th February for the last one. of average ues of current and we were Figure - 3 b shows the positions of waves, able to confirm the calibration numer- devices, which were organized as of the ical models, in particular in terms of described in Table - 1.

47 Messina Conference 2005

mean depth-averaged velocity, account through an additional sink Depending on the point, the velocities term in the r-h-s: were not as strong during the ebb flow OU F S D, v'U = — x — —— ' r. —+ U -g'Pq+ v(hD)+' yk U+ + I"IU (2) and the tide flow. Br hI, h i( 2.4, . A second series of measurements have been also performed in April in the In equation (2), U, h and h are respec- Brittany waters, in the area of "Les tively velocity vector, water depth and H eaux-de-Brehat", according to the surface elevation, while D is a diffusion- same protocol, but only in two points. dispersion tensor, g a Coriolis parameter and F and S bottom and wind 2.2 - The "farm-size" picture: model- friction. The last term represents the ling current-structure interactions effect of a converter on currents, Ds, As and CD being an equivalent diameter, The design of a complete marine cur- an equivalent horizontal surface and a rent converters farm requires the capa- drag coefficient, respectively. bility of predicting the main Practically, instead of using a mesh like interactions that can occur between those presented in Figure — 4.a, the converters and tidal currents. Indeed, proposed method works on the kind of not only the turbines work through cur- mesh showed in Figure — 4.b. rents, but they have also an influence on them in their vicinity. Effects of shear stress and diffusion can be responsible for a dramatic decrease of momentum in the converter's wake, These phenomena can be quite complex in a reai situation since tidal currents are unsteady and affected by bottom levels. As a consequence, a correct prediction and optimisation of the power which can be extracted from a turbine farm should be obtained through numerical modelling. Figure - 4: Two finite elements meshes We go on using the 2-D shallow-water used to computed the velocity field in equations (or Saint-Venant equations) the vicinity of a square structure. In 4, a which are suitable for modelling tides (up), the structure is meshed, while it is in coastal areas. In this context, one not in 4. b (down) problem is the numerical modelling of a converter: a local mesh refinement would require a high computational effort and would not allow to model the effect of turbines. Alternatively, one may consider that a structure acts on the flow as a strong shear area (Hervouet, 2003), In the depth-integrated momentum equation, the effect of such a shear stress will be taken into IVlessina Conference 2005

In this case, As will be the surface of 8 The abovementioned model was first triangles covering the bottom area tested in the telemac-2d code. The occupied by the converter, and the wake behind a square structure was drag term appearing at the end of modelled using both methods: Figure 5 equation (2) must be considered in this gives the distribution of velocity mag- only area. The drag coefficient CD can nitude with (a) a structure included into be estimated from experiments or the- the mesh and (b) the additional shear oretical investigations; for an accurate term in equation (2), showing that the modelling, one should mention the fact present approach provides fairly good that CD must depend on the velocity results. More precisely, Figure 6 shows direction, e. g. on the tidal stage. three velocity profiles computed with both methods.

el& eeee &ee&e 9 a„IB O~ 0 i~ LI LI Q~ 0„G~ 0e& D Le ~ I ~ I.e 0„0De& 00&& I„

I.e ~ I.I ~ e~ ~ ~ I

Figure — 5: Finite elements computation Figure - 6: Velocity profiles obtained of depth-averaged velocities behind a from the 2-D computation presented in square structure, using both meshes figures I and 2, at 150 m (a), 300 m (b) presented in figure 1. Comparison and 600 rn (c) downstream. between the case of a meshed struc- Comparison between the case with a ture (S.a, up) and the method consist- meshed structure (solid lines) and the ing of the additional shear term case with an additional shear force appearing in equation (1) (S.b, down). (dotted lines),

Starting from these promising results, the telemac-2d code is now used to compute the behaviour of velocities during a complete tidal cycle in the area of a converters farm. Figure 7 shows the wake effect due to 35 converters in a hypothetical area with complex bottom levels. Extensive works will provide recommendations to optimise the design of such a configuration, particularly the distance between converters.

49 Messina Conference 2005

steady re-circulation of water, lead ing to a modified time-averaged vertical profile of current (Figure 8).

o A dynamic part that considered the superposition of the oscillatory motion of water particles to the steady undisturbed conditions. Figure — 7: l3epth-averaged velocity field during flood in a potential area for 35 marine current turbines, simulated with 30 ~ Following the telemac-2d code. The wake effects 25 Wales are considered through equation ~ Opposing (1). 20 Wates 2 Undisturbed rn ~ - The "device-size" piicture: mod- 23 10 eiiing weve-current interactions

0 To be able to understand better and 0 0 05 10 )5 2, 0 optimise the instantaneous productibil- Current velocity intra) ity of a converter, it appears interesting - to refine the approach once again and Figure 8: simplified modeling of the reach the converter scale. Indeed, as velocity vertical profile reminded in 52. 1.2, the real structure of the water column is not that well The static part is shown to have a sig- known. nificant influence on the power output. In particular, how can the waves influ- Following waves tend to decrease it, ence the velocity in the water column while opposing waves tend to increase and then can lead to a fluctuation of it. the output power (with a short period) The dynamic part of the wave causes as well as a fluctuation of the mechan- large oscillations in power and thrust. ical loads on the turbines — this last This influences the "quality" of the point means dynamic phenomena and power (flicker) and causes fatigue to fatigue troubles for the different com- the drive train and to the support struc- ponents of the turbines ? To estimate ture. The detailed study shows that the amplitude of these fluctuations, the fluctuations are due to variations of the approach is twofold: numerical model- angle of attack under influence of wave ling and experimental measurements oscillatory motion, at sea, Though the wave influence is clearly - 2.3. 1 Numerical modelling illustrated, the models used in this study are not sufficient for a complete This study considers 2D vertical flows representation of the effect of waves with waves decomposed in 2 parts: and of the turbine behaviour. Phenomena such as the modification of o A static part characterised by a the mean current are still not well Messina Conference 2005

known. The empirical description of the Barf leur (point 4 — Figure 3a) the waves mean current profile under the influ- come from the North and North-West ence of waves, used in this study, is and have a lower frequency than the inaccurate and subsequent results Eastern waves. The waves' period times should be considered carefully. This is vary commonly between 2 and 9 sec- the reason why experimental data are onds, with differences linked to the useful. directions of waves,

2.3. 2 - Experimental measurements at Measurements also show that the sea stream direction has an influence on the characteristics of waves: when As presented in 5 2. 1.2, experimental waves and stream have the same direc- ' measurements were performed in tion, the height and the period-time of January 2005 in Nord-Cotentin in north waves generally decrease; they increase of France, While average values of cur- when stream direction is opposite to rent and waves measurements were waves direction and the waves energy performed to calibrate and to verify the is concentrated in low frequencies, numerical models (see 5 2, 1,2), some instantaneous values of current and Moreover we get some information free surface were measured too. This about the influence of surface waves will allow us to check the influence of on the vertical speed profile: in Barf leur, the waves on the water column; this according to our measurements, the section presents very preliminary wave height has very little effect on the results for these recent measurements. stream velocity profile. But it has some influence on the fluctuations (ampli- The position of the free surface was tudes and periods) of the tide stream: measured every 3 seconds. At the same such an influence can be observed time, the horizontal speed was meas- down to a depth of several meters, ured along the water column every 1 meter from the seabed to the free sur- The exploitation of such data related to face. These measurements allow to plot the influence of wave on stream is also the velocity for several wave positions. a important issue because the fluctua- In a first time we are going to see what tion of stream due to waves can lead to are the characteristics of waves near perturbations on output power. We will Barf leur and we will deal with interac- also focus on period times of the waves tion between waves and stream in a and compare the most common of second time. The measurements are them with the natural frequencies of based on the spectral density of ener- current turbines. The global data about gy in the waves because it is possible waves will have also to be used to make to deduce the height and frequency of a fatigue troubles' study. waves with such data. As expected the height of waves 3 - VERTICAL-AXIS TURBINES increase with the distance to but coast, EVALUATION: AN EDF-LEG I the measurements also gives other COLLABORATION general information about waves in different ways. For example in the East of To harvest marine current energy, two Messina Conference 2005

major kinds of water turbines exist: The hydrodynamic tunnel of LEGI horizontal-axis water turbines, and ver- (Figure — 9) has been modified to set up tical-axis water turbines or transversal- an experiment to be able to test and axis water turbines. Any potential compare different technologies of the end-user in the domain of MCE has to vertical axis water turbine. strengthen its own point of view on the various concepts and designs being developed before deciding to invest in The turbines are to be tested in a any. hydrodynamic tunnel which has been modified to accept one vertical axis To illustrate which kind of work is done water turbine. The turbine is placed in by EDF to get prepared and anticipate, a rectangular tunnel; three sides of the the development of suitable tools tunnel are in plexiglas, so that we are (experimental and numerical) ongoing able to observe the turbine on three on vertical-axis water turbines in the sides: on the lateral sides and the bot- LEG!, a French laboratory specialist in tom side of the tunnel. All the sensors hydraulics and turbo machinery. The for the experiment are mounted on the work described in this paper is co- top of the tunnel. financed by EDF and the French The tunnel shape was chosen to be Government agency ADEME. Indeed, able to study the turbine slipstream. different technologies of transversal- axis water turbines are developed in Different types of turbine are to be test- different places around the world, and ed in this tunnel, In a first step, it is EDF estimates tools are required to planned to test one Darrieus turbine compare such different technologies, (Darrieus, 1926), one Gorlov turbine between themselves and with horizon- (Gorlov, 1997) and one Achard turbine tal-axis water turbines, The tools being (Figure -10), Thus, the Kobold turbine developed at the LEGI are both experi- was not considered in this work, but mental and numerical, complementary. the principles of analysis could apply the same way. 3.I - The experiment

3. 1. 1 - Description of the experiment

'Qj pj's' 0 1 4&6

Figure - 10: Photo of Darrieus turbine Figure - 9: Photo of the tunnel (left) and drawings of Gorlov (right) and IVIessina Conference 2005

Achard4 (down) turbines which will be This measuring system includes rails tested in the tunnel of the LEGI (source which support the system in the meas- Mile Kusulja) (manufacturer SERAS, a ured force directions. So, the system is department of CNRS). able to move slightly in order to make the force sensors react. Some parameters of the flow like pressure and dis- charge are also measured,

3.2 — Aims

The experiment is a good source of information about the performances of different types of turbines, From the rneasurernents, we will deduce the extracted power. We will also determine the characteristic curve, valid for all the turbines in similarity with the tested turbine. According to previous numerical stud- 3. 1.2 - Measurements ies in LEGI and to experimental studies in other laboratories, we should find a To be able to determine the turbine power coefficient maximal for a tip performances, of course, we will meas- speed ratio, 2, between 1 and 3. The ure the rotation speed of the machine tip speed ratio which characterised the and the torque it produces. Moreover, rotational speed of the turbine with we will measure forces applying on the regards to the flow speed, is defined by turbine, Forces are measured thanks to the following equation: a system made in LEGI. The directions of the measured forces are: Rm (3) V 0 the one parallel to the flow with R turbine radius; Arotational speed o the one parallel to the rotation axis of the turbine; V flow speed in the upper 0 and the one perpendicular to the 2 waters far away from the turbine. other measured force directions. In our experiment, there is also a will to

Axis I fnrce learn about the forces applied to the whole turbine. The knowledge of the forces will be precious information to Flow Dray calculate the size of water turbine farm structure and of farm anchorage. For these calculations, it is good to have I' rc per ndioular to the average values but also the peak th- values. We also expect to measure the variabil- — Figure 10: Schema of the measured ity of torque and forces on one revolu- forces tion. Indeed, because of the rotational

53 Messina Conference 2005

movement and dynamic phenomena However, they predict more precisely like dynamic stall, the forces on blades efforts but need more calculating time. are varying a lot during one turbine revolution, As there are a small number For the models based on the momen- of blades, the variations are not totally tum theory, three types can be distin- averaged for the whole turbine. The guished: forces on the turbine are difficult to calculate: the experiment is useful to o Single streamtube model: the whole study these forces. The knowledge of turbine is inside a single streamtube, the forces variation will be useful to It is modelled by one receptor disk. study the fatigue of the turbine and axes material. o Multiple streamtubes model: the Some comparisons between turbine flow going through the turbine is technologies can be made thanks to split into several streamtubes, The the characteristic curves and all the turbine is modelled by one receptor force measurements. So, this experi- disk. the dif- ment is a good tool to compare o Double-multiple streamtubes model ferent turbine types on the : the flow going through the turbine performances but also on the forces is split into several stream tubes. applied to material and anchoring The turbine is modelled by two problems. receptor disks: the flow passing through them successively. 3.3 - The numerical aspects

will information The experiment give These models, and some improve- only for few turbines, To be able to esti- ments, have been studied notably by mate performances of a wider range of Sandia laboratories (Strickland, 1975) turbines, a little code is developed. The and Ion PARASCHIVOIU (Paraschivoiu, code results will be compared to the 2002). At present, LEGI work on these experiment results to estimate the three models, code efficiency and then to improve it. The code is based on the numerical To improve these numerical models, model used for vertical-axis wind tur- one must consider dynamic stall, down- bines. Two different kinds of models stream blades going through upstream were studied; blade wake, streamline divergence, relative flow streamlines o Models based on the momentum curvature. . .These improvements must theory be important to adapt these models to o Models based on vortex theories vertical-axis water turbines,

LEGI has studied models based on the momentum theory. All these models are inspired by Glauert work on pro- This paper presented various activities, peller, To estimate the global perform- mostly ongoing and preliminary — ded- ances of the turbines, vortex models do icated to "fluid" modelling and testing, not give better results than models at EDF RRD and its partner LEG!, to based on the momentum theory. improve the knowledge related to Messina Conference 2005

Marine Current Energy in France. Une source d'energie renouvelable possible les hydrauliennes, Revue de Enough has been done to highlight the I'Energie, Les editions techniques et cornplementarities of numerical and economiques, n'546, May 2003. experimental approaches, both on the currents' studies and the harnessing [6] Achard, J.L. et Maitre, T. 2004, devices' analysis. From the desk to the Turbomachine hydraulique — French sea, going through the laboratory, and patent FR 04 50209, 4th February 2004. vice-versa. . . [7] Strickland, The Darrieus turbine A lot remains to be done in the years : a performance prediction model using to come for the first devices to produce multiple streamtubes, Sandia laborato- first kWh, in structural mechanics, elec- ries, October 1975, SAND75-0431 trotechnics, electronics. . . We hope that studies like the ones presented in this [8] lon Paraschivoiu, Wind Turbine paper will contribute to achieve such Design: with emphasis on Darrieus an ambitious target, even if, due to the concept, Polytechnic International recent start of these studies, most Press, 2002, results are still to come, They will be shared on time. GLOSSARY

'Tidal stream energy' is synonymous REFERENCES of and alternatively used for 'marine current energy (MCE)' in this paper, [1] CENEX: Marine Currents Energy Extraction: Resource Assessment, Final Report, EU-JOULE contract JOU2-CT93- ' AWAC = "Acoustic Wave And Current" 0355, (Tecnomare, ENEL, IT Power, ADP and Aquadopp = Acoustic Ponte di Archimede, University of Doppler Profilers, by Nortek. Patras) 1995 Doppler principle: change of fre- [2] Hervouet, J.M. (2003), quency due to the velocity of a source Hydrodynamique des ecoulements a or receptor surface libre. Modelisation numerique avec la methode des elements finis, ' Jean-Luc Achard, Professor, CNRS, Presse des Ponts et Chaussees, Paris. LEGI. LEGI has actually its own MCE project, named "HARVEST" ans aiming [3] G. J-M. Darrieus, Turbine a axe at the development of this ACHARD de rotation transversal a la direction du device (Maitre 8 Achard, 2003) (Achard courant, Brevet d' invention, n'604. 390, & Maitre, 2004). 3 mai 1926. [4] Alexander M, Gorlov, Helical turbine assembly operable under mul- tidirectional fluid flow for power and propulsion systems, United States Patents, n'5, 642, 984, 1st July 1997.

[5] Thierry Maitre, Jean-Luc Achard,

55 Messina Conference 2005

Fuel Cell Microcars for Small Sicilian islands

V. Antonucci, F.V. Matera, L. Andaloro, G. Dispenza, M. Ferrara

ABSTRACT 1 - INTRODUCTION

CNR-lTAE is developing hydrogen fuel The forecast growth in fuel consump- cell microcars for small Sicilian islands. tion by road transport has been identi- These vehicles are lightweight and fied as a major policy challenge in the zero-emission. Regenerative braking European Union (EU). This can be and advanced power electronics make explained by concerns about fuel con- these vehicles very efficient. Moreover, sumption-related emissions of green- to achieve a very easy-to-use technolo- houses gases (GHG), and the increasing gy, a very simple interface between dependence on imported oil, driver and the system is under develop- In this context, it is generally agreed ment, including fault-recovery strate- that actions need to be taken to curb gies and GPS positioning for car-rental GHG emissions. Thus, strong interest fleets. The project includes a hydrogen has arisen in the last decade for fuel refuelling station powered by renew- cell-powered vehicles. able energy, in order to make the over- Polymer electrolyte membrane fuel all system self-sufficient, as well as to cells (PEMFC) are the most promising test the technology and increase pub- power sources in the near future for lic awareness toward clean energy residential, mobile and automobile sources. applications. They offer the potential of compactness, lightweight, high power Keyword: fleet, fuel cell, powertrain, density, easy system management and regenerative braking, ZEV (zero emis- low temperature . operation. They are sion vehicle) actually also considered the best solution for automotive applications. This is due to criteria related to large flexibili-

57 Messina Conference 2005

ty, fast start-up, small dimensions, high height above sea level (Table - 1, right). efficiency and low working temperature; this latter characteristic is impor- I!It!i& tant for safety aspects. PEMFCs use QU!!i

il!!0.10 highly conducting electrodes made of i'Flj , li iilC terminal of i graphite, which form the Nii[ljtytl each cell and separate adjacent cells in I I&Nil!i the stack. The electrodes are grooved to allow easy passage of the gases to I! the 'surface of action' while also main- , %f I!i!i!i taining electrical contact with the electrolyte-catalyst-gas interface. At the anode, hydrogen is catalytically dissoci- ated to leave hydrogen ions. An exter- Figure - 1: Sicilian islands layout nal circuit conducts electrons while the positive ions (protons) migrate through Often, during wintertime, they remain the electrolytic membrane to the cath- isolated due to atmospheric and sea ode. There they combine, again under conditions. Main supplies, such as action of a catalyst, with oxygen and potable water and fue), are provided by electrons returning from the external bulk carriers from the mainland, circuit to form water. The state-of-the- excepting some islands where drinking art catalysts for both electrodes are water is produced by desalination of based on carbon-supported platinum; sea water. Power is locally produced by the electrolyte is an ion-conducting diesel generators connected to the perfluorosulfonic membrane. Stacks are local grid. obtained by connecting a number of The main human activities are agricul- cells in series while gas distributors ture, with valuable typical products, (bipolar plates) accomplish the func- fishing and tourism, Although tourism tion of feeding the fuel to the anode infrastructures are not always very well side and air (or oxygen) to the cathode. developed and organized, population grows roughly S times during summer holidays. in recent years small islands have Vehicles are generally not allowed dur- attracted major attention, with several ing summer for non residents; howev- European and national programs er, the most used means of transport addressed to the development and are gasoline cars or smoky and noisy protection of these areas. Small Sicilian two/four strokes vehicles. islands, recently declared as mankind heritage, include large portions of protected areas; thus, they represent a good opportunity for sustainable development solutions. The 14 Sicilian islands (Figure — 1), organized in 8 municipalities, extend from 3 to 38 sq. km, with a significant Messina Conference 2005

Weight asl Island Municipality District Area, sq. km (min-max), m Lampedusa 20.2 0-193 Lampedusa AG Linosa and e Linosa 5.3 and 1,2 0-195 Lampione Lipari 37.6 0-924 Vulcano 21 0-500 Strmboli and 12.6 0-926 Strombolicchio

Filicudi Lipari ME 0-7? 5 Alicudi 5.2 0-675 Panarea and surrounding 3.0 0-421 rocks S. MArina di Salina 0-963 Salima Malfa 27 0-860

Leni 0-962

Ustica Ustica PA 8.65 0-248 Favignana 0-686 Favignana TP Marettimo 12 0-984

Leva nzo 10 0-278

Pa nte lie ria Pa ntelleria TP 83 0-836

Table — 1: Area and height abave sea preferred from an environmental point level of small Sicilian islands. of view, but it has strong limitations due to short operating range and long 3 - FUEL CELL VS. BATTERY recharging time. A disadvantage of current Electric Veichle technologies is that the In small islands, typical use of vehicles energy supply required to vehicle mainly include rent car for tourists, taxi power the needs to be stored and shuttle service, with no special as electricity on board the vehicle, The batteries that are this pur- requirements in terms of speed, These needed for vehicles must also be able to overcome pose are expensive, have a short and have a limited steep slopes even at full load, and recharging cycle, should be noiseless and smokeless. storage capacity, The range of an elec- The electric traction, today based on tric vehicle is therefore much smaller than for vehicles with conventional batteries, is currently fueled gaseous or liquid energy carriers. Batteries currently have a recharging time of approxi-

59 Messina Conference 2005

mately 6 to 8 hours, which makes these 4 - MICORQCAR DESIIGN cars less suitable for intensive use. Replacing empty batteries with At present, the automotive market is charged batteries is at present not con- investing in microcars. Microcars in sidered a serious option, mainly Europe are about 250, 000 (more than because the batteries are too heavy. 30,000 in Italy) and, in 2003, 10,000 microcars have been sold vs. 8.500 in Although several models of battery- 2002 (18~/a increase). They are an alter- powered light vehicles are commercial- native model of mobility thanks to their ly available, most of them are not characteristics (small size, limited suitable for these applications. Very speed, reduced consumption, easy steep slopes require an high torque driving). motor (or a motor coupled with a reduction gearbox) and a large amount Microcars can be advantageously used of energy stored on board. In the case in areas where land morphology (nar- of battery-powered vehicles, this would row roads, steep slopes) or other require a large number of batteries restrictions (limited traffic areas) could installed onboard with strong limita- cause transportation problems. Hence, tions in terms of weight and volume. small islands and natural parks can represent a real laboratory for testing and The increased weight would dramati- development of these vehicles, cally reduce acceleration performance CNR-ITAE is developing this new gener- of the vehicle and increase the need of ation of electric vehicles, in collabora- an oversized braking system. Also tion with industrial partners involved in goods or luggage on-board storage fuel cell microcars production. The goal capacity would be sacrificed. Moreover, is to realize a transportation system battery disposal would rapidly become able to solve the problems connected a problem without an adequate and to individual mobility in protected efficient policy of waste battery man- areas with efficient, noiseless and non agement. polluting vehicles, as well as to intro- duce an easy-to-use technology orient- Recharge is another critical issue for ed to tourism and car rental, battery powered vehicles. The long recharge time, requires an adequate The vehicles under development vehicles fleet management. Although include a hydrogen fuel cells generator, some alternative solutions have been a storage system for energy recovery proposed, such as battery-pack turn- and lightweight compressed hydrogen over or on-board power generation tank. An advanced interface between with small reciprocating (gasoline or the fuel cell system and the electric diesel) engines, none of these seems to motor allows regenerative braking and really meet the need of a clean and effi- energy management, with consequent cient vehicles fleet; moreover, the prob- high overall efficiency. A further advan- lem of battery disposal would remain tage, with respect to battery-powered to be solved. vehicles, consists in the fast refuelling of the fuel cell system (few minutes vs. up to eight hours). Moreover, exhaust

60 Messina Conference 2005

battery management would not more speed gearbox to face steep slopes at be an issue. full load. Three models are being The vehicles are provided with GPS designed, having two or four seats and positioning and diagnostic system; this, different volume for storage tanks and in case of fault, reports the vehicle sta- luggage compartment, in order to tus, the type of fault (as indicated by meet various requirements (hotel shut- sensors) and the vehicle position to a tle, car rental, goods transport). remote unit which is able to interact with the driver for assistance, The power requirements of the vehicle can significantly differ, depending on Powertrain configuration specific needs (weight to be transport- ed, driving cycles, etc. ). Thus, some The vehicles are designed to meet the degree of hybridization is needed, italian regulation on "light quadricy- cles". The principal requirements are a Two powertrain configurations ( paral- maximum power of 4 kW, speed limit- lel and series) are depicted in Figure- ed to 45 km/h and maximum weight of 3 and Figure — 4 ( next page) 350 kg for the whole vehicle excluding battery pack. Figure — 2 depicts a gen- In the parallel hybrid powertrain eral scheme of the power train, (Figure - 3) most of the power for the electric motor is produced by the fuel cell. Moreover, II the fuel ceil provides COOL'ING sySTEM II to I I recharge batteries when I I I II I required, in this case, bat- I I ) teries can assist the fuel cell 14 OIR I GK» when peak power is neces- I I I I I I sary (acceleration, climb, I I I I I I etc. The battery acts I ) pack I I also as power storage in it EIIERGV II FLIEL WORDAGE I I I MGDL'LE the regenerative braking. I II I In I the series configuration I I (Figure — power is pro- I POWER MOOULE I I 4), I I oo,"IER VGDLL~ vided by the battery. The USER CGNTROL I CGII COLS WI4D INTERFACE I fuel cell works only as a for I \ I l the battery charger; this system is suit- PQWERTRAIN particularly

DPTALGGG444G LOAD SIMUL4VIGI I able for urban driving iNLI SVPCKVI'GK KPKGIVPK UII IT cycle. Logicol conneCtion

Pttyslcet connection Where steep slopes are present, a large amount of energy can - Figure 2: General scheme be recovered during downhill driving of the powertrain (regenerative braking). This recovered energy is one of the most important The powertrain is designed with a two factors that increase the overall effi- Messina Conference 2005

ciency of the vehicle, as it affects the electrolysis can be an interesting primary energy consumed (hydrogen) option. per km (i.e, J/km). Nowadays, current fuel cell vehicles make use of centralized refuelling sta- Ballet@ ol tions. Buses and rnicrocar fleets are par- n tpe1 c 1111a c1 1 or ticularly attractive as an entry market Regeneration. 1 echarRe hattely tehcn needed for fuel cell vehicles, as they are central- Auctllath maintained, Jecicet Ac»st Fnel Cell ly garaged, refuelled and nhen neecied Near term demonstration projects will help to disclose most stringent techni- Fnel Cell h lotol DCiDE hlcenel cal issues for different of fuel cell atacl' conc types vehicles. — Pon et' Figure 5 depicts a general view of the teqntrentent envisaged system.

Figure — 3: Fuel ceil parallel hybrid sTATIOX'ARY power train configuration

Ancillary dectcea ENERGY

ELECTltJC ENERGY

Encl Cell DClD Datterr inrel tel hlolor track conc

H RFCtRAGF

Figure — 4: Fuel cell series hybrid power Figure - 5: General scheme of the envis- train configuration aged system

The energy produced by photovoltaic panels and wind generators can be par- tially used for residential purposes; the excess could be transformed, by elec- In recent years, several projects for the trolysis, in hydrogen which is stored exploitation of renewable energy have then used for the fuel cell-powered been set to assess the potentiality and and vehicles. flexibility of such sources in these islands. In such places, the cost of elec- - trical energy is typically very high and 6 COMIClLUSIONS often the surplus cost is paid by public administration. Production of hydrogen fVlain goal of this project is to design can represent a possible solution to and test a small fleet of fuel-cell hybrid store the excess energy via renewable electric vehicles under the cited envi- sources (solar, wind, etc. ). Since ronmental and background limitations. methane reforming is not a viable solu- The protection of environment in tion in isolated localities, due to the places such as Sicilian Minor Islands is lack of distribution infrastructure, water a priority. In this view, the proposed Messina Conference 2005

application justifies the actual higher Power Sources 79 1999 cost of fuel cell vehicles. Moreover, the close integration between hydrogen [7] Gert Jan Kooprnan, Long-term production by renewable energy Challeges for inland transport in the sources and zero emission vehicles European Union: 1997-2010, Energy gives today the opportunity to test, Policy, vol 25, 1997 under favourable circumstances, a [8] J. Larminie, A, Dicks, Fuel Cell energy technology that is considered Systems Explained, Second Edition- one of the most promising in the long Wiley, 2003 term. [9] Ludmitla Schlecht, Competition and alliances in fuel cell power train devel- REFERENCES opment, Hydrogen Energy 28 (2003) 71 7-723 [1] A. Armstrong, Fuel Cell Technology [10] J. Linssen, Th, Grube, 8. Hoehlein, Conference, University of California, M, Walbeck, Full fuel cycles and market Davis, USA, 30-3'1 March 1999 potentials of future passenger car propulsion systems, Hydrogen Energy [2] B.D. McNicols, D. A.J. Randb, K. R, 28 (2003) 735-741 Williamsc, Fuel cells for road transportation purposes — yes or no?, Power [11] Mare W. Melaina, Initiating hydro- Sources 100 (2001) 47-59 gen infrastructures: preliminary analysis [3] Steven G. Chaik, JoAnn Mitliken, of a sufficient number of initial hydro- James F. Miller, S.R. Venkateswaran, The gen stations in the US, Hydrogen US Department of Energy, investing in Energy 28 (2003) 743-755 Clean Ttransport, Power Sources 71 (1998) 26-35 [12] K.-A. Adamson, An examination of consumer demand in the secondary [4] Q. Robert, Riley Enterprises, electric niche market for fuel cell vehicles in and hybrid vehicles: an overview of the Europe, Hydrogen Energy 28 (2003) benefits, challenges, and technolo- 771-780 gies, Technical Paper, June, 1999 [13] S, Srinivasan, R, Mosdale, P. [5] Koen Frenkena„Marko Hekkertb, Stevens, C. Yang, Fuel cells: reaching Per Godfroijc, R&D portfolios in envi- the era of clean and efficient power ronmentally friendly automotive generation in the twenty-first century, propulsion: Variety, competition and Annu. Rev. Energy Environ, 24 (1999) policy implications, Technological 281-328. Forecasting & Social Change 71 (2004) [14] M. De Francesco, E. Arato, Start-up [6] Joan M. Ogden, Margaret M. analysis for automotive PEM fuel cell Steinbugler, Thomas G. Kreutz, A com- systems, Journal of Power Sources 108 parison of hydrogen, methanol and (2002) 41-52 gasoline as fuels for fuel cell vehicles: implications for vehicle design and infrastructure development, Journal of

63 Nlessina Conference 2005

AUTHORS

Vincenxo ANTONUCCI, Dr. , Fue! Cell Research Manager, CNR- ITAE, Salita S. Lucia sopra Contesse n' 5, +39-090-624.234, f a x: + 3 9 - 0 9 0- 624. 247, antonucci@itae. cnr. it

Fabio MATERA, Dr. , Automotive Fuel Cell Systems Dep. — Researcher, CNR — ITAE, Salita S. Lucia sopra Contesse n' 5, +39-090-624.236, fax: +39-090-624.24?, fabio. matera@itae, cnr. it

laura ANDALORO, Dr. , Automotive Fuel Cell Systems Dep. — Researcher, CNR - lTAE, Salita S. Lucia sopra Contesse n' 5, +39-090-624.235, fax; +39-090-624.247, la ura. and a toro@itae. cnr. it

Giorgio DISPENZA, Dr. , Automotive Fuel Cell Systems Dep. - Researcher, CNR — ITAE, Saiita S. Lucia sopra Contesse n' 5, +39-090-624.235, fax:+39-090-624.247, dispenzaOitae. cnr. it

Marco FERRARO, Dr. , Stationary Fuel Cell Systems Dep. Researcher, CNR - ITAE, Salita S. Lucia sopra Contesse n' 5, +39-090-624.231, fax: +39090624247, ferraro@itae. cnr. it Messina Canference 2005

The European Commission's activities in support of Renewable Energies, in particular Ocean Energy

Komninos DIAMANTARAS

Despite attempts to decoupte econom- of OECD countries going down from ic growth from energy consumption (in more than 50% to about 44%}. progress to some extent in industri- alised countries), energy is currently an Energy is therefore at the core of glob- important component of growth. As al concerns, as exemplified by the events have recently demonstrated, an Climate Change Convention and Kyoto increase in the cost of energy immedi- Protocol aiming to reduce greenhouse ately translates into a decrease of GNP gas emissions through cleaner and growth and in the well being of more efficient use of traditional ener- European citizens. Energy price volatil- gies, and introducing Renewable ity of the kind currently being experi- Energy Sources (RES) as alternatives to enced also has a determining effect on fossil fuels, and the Johannesburg economic growth and investment deci- Renewable Energy Coalition Joint sions. Europe therefore requires a sta- Declaration "The Way Forward on ble, secure and affordable energy Renewable Energy", and in a more gen- supply. eral context the Goteborg and Lisbon Strategies on Sustainable Growth of energy consumption at Development, world level is due to increase sharply in the years to come especially with the I will therefore present the major ener- somewhat unforeseen economic boom gy research policies in place and will of demographic giants such as China refer to the research activities in the and India (conservative figures indicate field of renewable energies with a progression from 10 000 Mtoe in 2001 emphasis on the sector of ocean ener- to 16 300 Mtoe in 2030, with the share gy, a short summary of which follows. Messina Conference 2005

Over the last twenty years, the the areas of shoreline and offshore European Union financed ocean, wave wave energy devices, of tidal current and tidal energy research and develop- turbines and of salinity gradient sys- ers. In total, twenty-nine projects have tems. Salinity gradient systems are a been awarded to research and devel- recent development and could be opment in the three main areas, of deployed in many European river estu- which four projects under FP5, one aries. under FP6, while one is still under negotiation, In FP5 the EC contribution The flagship prototypes developed was 4.54 M, and in FP6 so far 4, 8 M, with EU financial support are: while 2. 5 M are still in the contract preparation phase. The cumulated EC o Shoreline Wave Energy: two de- contribution during the last fifteen mostrators of 400kW each, one years sums to about 23 M with a total on the island of Pico, Azores, eligible cost of the order of 48 M. and one on the island of Islay, Scotland (FP4 projects) Increasing RKD funding is critical to o Tidal current turbine: one proto- advancing the development of ocean type of 300kW, (FP4 project) energy systems. Ocean energy technologies must solve two major prob- 0 Kobold marine current turbine: lems concurrently: proving the energy one prototype of 12 kW (FP4 conversion potential and overcoming a pl OJeci), ancl very high technical risk from a harsh environment. No other energy technol- 0 Offshore Wave Energy: one 1:4 ogy has had to face such demands. prototype of 20kW, — known as When their prototype, deploying Wave Dragon (FP5 project) device developers are confronted with the possibility of losing five years of development and investment in few The following projects under FP6 have hours time, Furthermore, the majority started recently and are at the imple- of the developers are SMEs for whom mentation phase: such a loss can be overwhelming. Additional RB, D funding would help to The Coordinated Action on mitigate the substantial technical risk Ocean Energy systems, CA-OE — faced by device developers daring to with 1,5 M C funding harness the energy of the marine envi- The SEEWEC project (point ronment. absorbers, offshore) with - 2, 3 M EC funding Ocean energy systems cover a wide The WAVESSG project (overtop- range of applications that can be ping type, suitable for breakwa- deployed on the shoreline and off- ters) with - 1,0 M EC funding, shore. Technology is emerging to allow and large scale demonstration projects, At present, very few demonstration proto- The WAVE DRAGON 4MW (4MW types exist and they are mainly in floating overtopping type, off- Europe. The research activities cover shore) with total cost of - 14,7 Nlessina Conference 2005

M out of which —2, 47 M EC LIMPET tlslay, Scotland): funding„ http: I/www. wavegen. co.uk/what we offe r limpet i slay. htm According to the IEA's the Renewables Internationa! Information report, during 2003 more Energy Agency: //www. iea-oceans. org/index1, htm than 572 GWh of electricity were gen- http: erated from wave, tide and ocean motion. The bulk of the electricity pro- WAVETRAIN (Illarie Curie Actions, FP6): ht t wave ra duced was generated in France p://www. t in, info/ through the 240 MW installed capacity related to the project. Canada was the second high- b) European Commission's activities and est electricity producer with about - 33 events: Gwh, EUROPA: http: I/europa. eu. int/comm/dgs/resea rc It is also worth noting that since h/index en, html October 2001 the European Commission participates and follows eu. through the Implementing Agreement http://europa. int/comm/research/en ergy/index en. html on Ocean Energy Systems the latest developments at international level I/europa, while promoting the research, develop- http: eu. int/comm/energy/ind ex en. html ment, information exchange and demonstration of Ocean Energy INCO and Marie Curie: Technologies. http: //www. cordis. lu/inco/home. html For additional information the reader http: eu. int/comm/research/fp can also refer to the following web- I/euro pa. sites; 6/ma riec uric-action s/indexhtm en. htrnl a) related to Ocean Energy: CORDIS:

h t cordi I European Wave Energy Atlas: tp://www. s. u/fp6/ http:I/www. ineti. pt/proj/weratlas/ RENEWS newsletter: http:I/europa. eu. int/comm/research/en Coordinated Action on Ocean Energy: http://www. ca-oe. net ergy/pdf/renews3. pdf

Wave-Net: Information days and similar events, con- http: //www. wave-energy. net/index3 .htm ferences: http: I/europa. eu. int/comm/research/en events/action/arti- Wave Energy Centre (P): ergy/gp/gp cle 2790 en. htm http: I/www, wave-energy-centre, orgl

WAVE DRAGON: http:I/www. wavedragon. net/

67 ~ ~ ~ ~ ~

UNIOO

UNITED NATIONS INDUSTRIAL DEVELOPMENT ORGANIZATION Vienna International Centre, P. O, Box 3oo, 14oo Vienna, Austria Telephone: (+43-1) 26o26-o, Fax: (+43-1) 26926-69 E-mail: unido@unido. org, Internet: http: //www. unido. org