Final Report Summary - HAWE (High Altitude Wind Energy)
Executive Summary:
This report describes a technology for harnessing energy from high altitude wind through a pumping cycle, in a vertical trajectory, executed by a hybrid lighter-than-air tethered rotating cylinder, which generates dynamic lift through the Magnus effect.
The pursuit of high altitude wind energy has led to a number of different systems and configurations which try to harness energy from the greater power densities available at altitudes higher than those used by conventional wind turbines. The HAWE consortium, comprised of 2 research institutes, 3 large private companies, 1 air traffic management authority and a Portuguese innovation-led SME coordinator, has developed, over the last four years, a system based on a hybrid airborne platform, which through the Magnus effect, converts wind power to force and speed (mechanical power) on a tether cable; as the platform executes a pumping cycle trajectory, mechanical power is converted into electricity, at a ground station. The concept is based on a PCT patent granted to Omnidea in 2010.
After initial trials with an airfoil in 2007, a rotating cylinder was selected as the lift force source for the Airborne Module, allowing the team to take advantage of the greater values of aerodynamic lift and drag available from the Magnus effect. Since 2011 the concept has been refined and further developed by the consortium, focusing on cycle modelling and optimization, as well as full system design and study. A 16m long by 2.6m wide, cylinder shaped, scale prototype, estimated to generate 3kN of force under wind speeds of 7m/s, was built at Omnidea and tested extensively at Ota Air Force base, where proof-of-concept was obtained, under operating restrictions, on Q3 2014.
This wind energy producing device is aimed at lowering the current wind generators benchmark costs (typically twice as costly to produce electricity vs. conventional fossil fuel sources) by aiming to make it competitive with coal derived electricity costs. Although the small-scale prototype could not confirm the significant cost reductions envisaged, the economic feasibility study conducted confirmed that, under the right operation parameters, the technology can be competitive with technologically mature wind turbines. Further to this, the aeronautical safety operation assessment evidences that, despite current air traffic regulations not being adapted to permanently moored/cycle operating platforms, delimited and identified demonstration parks can be installed if outside the influence area of airports and large cities. A detailed study for off-shore operation demonstrated the suitability of the concept to be installed, in deep waters and on “shore distant” locations, likely through the use of adapted tension leg platforms. This development also saw the issue of a patent application within the period of the project.
During the development stage, challenges where identified and solved. Still, the results achieved, at sub-system testing, uncovered components where even the current technological state-of-the-art must evolve: the multi-layer, multi-function Dyneema tether cable must improve to achieve a greater bending fatigue resistance, while different production techniques must be experimented with, to allow insert energy transmission copper conductors in a structural cable subject to millions of cycles per year.
Project Context and Objectives:
Currently all produced wind energy comes from Horizontal Axis Wind Generators, known colloquially as wind turbines. Modern wind turbines use lift force, generated as the incoming air is deflected around the turbine blades, to rotate the blades around the housing hub. Turbines incorporate features such as pitch mechanisms and yaw drives to increase extractable wind power, but are subject to the Betz limit (from the afore mentioned wind deflection) which limits maximum extractable power to 59% (16/27) of theoretically available wind power. But, as the turbine operating efficiency varies between 60 and 75%, in reality only around 40% of the wind power density available is translated into electrical power.
Commercial turbines on the market range from 0,5 to 6MW and start rotating at wind speeds of 2,5m/s, only being shut down if wind exceeds 25m/s, even though their rated (nominal) power is attained with wind speeds around 12-13m/s. Nevertheless, since constant winds at hub height are usually much lower, at around 6,5 to 8m/s, power production is only around 200 (minimum on shore) to 400kW (maximum off-shore) per installed MW. In June 2014 over 336GW of wind power were installed worldwide, and estimates for power generation, for 2014, should be close to 1 PWh (1 Peta Watthour) of electricity, representing 4-5% of all electricity used worldwide .
In trying to increase wind power production, the current wind tower technology has been aiming for ever higher hub altitudes, due to the existence of a greater power density in wind (i.e. wind blows stronger and more constant) but also because this allows for longer blades (i.e. a greater swept area). This poses an increased challenge in the towers, as the mass of these increases substantially with height. The engineering challenges that this quest presents demand an enormous financial investment in R&D and construction.
Airbone Wind Energy harnessing devices, differ from conventional wind turbines in that they extract energy from wind without the need for a turbine tower. Instead this is usually replaced by a tether cable connecting an airborne installed device to a ground station. The airborne installed device can extract wind energy:
• directly, doing the conversion to electricity in the airborne device, sending electricity through the tether cable to the ground station, or
• it can produce mechanical power which is then converted to electricity at the ground station, from the movement-force “pair” reaching the ground station through the tether
In what respects to the airborne device, it might produce energy:
• constantly (i.e. “always on” or always producing) or
• intermittently (for example through the use of a pumping cycle, similar to reciprocating engines, where a phase of energy production is alternated with energy spending to bring the cycle to its original, i.e. “start” position)
In what regards to the construction nature and principle of operation, the ABM can be:
• an aerostat, relying solely on lighter-than-air technology to stay aloft;
• an aircraft, being heavier than air and employing aerodynamic lift to stay aloft;
• or a hybrid, employing both technologies simultaneously.
Airborne Wind Technology represents a derivation from the conventional wind power development, with a multiplicity of technological options. Airborne wind harnessing systems have multiple advantages over conventional wind turbines:
• no need for a heavy and expensive turbine tower, which also means foundations are much less problematic;
• possibility to install the electricity generating equipment on the ground;
• much higher wind energy, from the higher velocities at altitudes above 1-2km, means there is no need to focus exclusively in optimization and efficiency;
• possibility to operate at different altitudes to increase capacity factor to 60-70%.
Additionally, airborne wind energy, aiming to attain high altitudes (above 1000m) is no longer influenced by the orography of the terrain over which it is installed, but rather subject to the availability of winds in the lower part of the troposphere. Potential seems to exist for significative cost reduction in electricity from wind energy, which will bring it in line with electricity generation from non-renewable sources (inc. coal). Nevertheless, as of 2013, apart from some demonstration parks, no such system was in commercial operation.
The HAWE project, was a 42 month, R&D centric technology development project, aimed at providing proof-of-concept and TRL increase of a hybrid airborne platform operating according to a pumping cycle arrangement to extract wind power and convert into electricity at a ground station, aiming to harness clean and renewable energy from high altitude wind.
The technology investigated consists of a lift generating airborne module (ABM) comprised of a rotating cylinder that is connected to a ground station by a tether cable, operating in a two phase cycle. During the power production phase the airborne module generates lift through the Magnus effect of a rotating cylinder, pulling up on the tether cable and converting the lift force into electricity at the ground station through an electrical generator. When the ABM reaches its limit of operation, the recovery phase begins as the rotation of the cylinder is stopped and the cable is reeled back to its initial position completing a cycle. The design and development of the three main components of the system, the ABM, cable and ground station involves a complex process with multiple interdependencies calling for several iterations to arrive at the design specifications needed for the development of the various components. The initial design capacity selected for the proof-of-concept demonstrator was a unit of 50kW, but for pratical reason this ended up being scaled down to a 10kW unit, in order to minimize time wasting efforts and focus on the development sub-system critical components.
The intermediate objectives were, therefore, to undertake an iterative study, design and implementation of sub-system critical components which could be individually tested before final assembly into the system (i.e. demonstration prototype). At sub-system level, the specific objectives were to design, manufacture and test:
(i) an ABM with the required aerodynamic characteristics, using inexpensive materials to provide a strong, gas-tight envelope that is UV resistant and capable of operating in typical wind environments. The system should have a low cut-in speed (1-2m/s) and a high cut out speed (20-25m/s) operating at altitudes greater than 200m;
(ii) a ground station consisting of a combined traction winch and electrical generator combining very high cable velocities with significant heavy loads and extremely long operating life spans, to perform a pumping cycle of the ABM for electrical power production, as well as the ground mechatronics layout and control software to optimize the operating strategy;
(iii) a non-metallic structural tether cable to secure the ABM, ensure data and electricity transfer between ABM and ground station as well as supply buoyancy gas to the ABM, whilst being able to withstand the stresses and pulling forces experienced in the operating cycles;
(iv) interfaces between the cable and the ABM consisting of swivels (one in the cable to avoid excessive cable twisting and another on the cable to winch interface, to always ensure the ability to transmit gas, electricity and data to the ABM.
HAWE was organized in 8 work packages plus a WP for management (WP90). WP 10 focused on the ABM design and construction, aerodynamics and main cycle simulation. WP 20, was essentially control related, featuring the layout of hardware for ABM control definition and selection of individual components as well as cycle optimization and control.
Cable design, including layout of the multifunctionalities (gas transfer, electrical power upload to ABM and data transfer and through cable communication to ABM), construction, testing and iteration, plus possible “large scale” manufacturing optimizations was the focus of WP30. Ground station development, from a design/construction point of view, was the main aim of WP50, which focused mainly on the winch/generator hardware part, plus the ground station control module. It run in parallel with WP60 which focused on energy storage and drive arrangement optimization at ground station, as well as components selection and control strategies for the control cabinet. WP40 centred on subjects related to more than a single module (whether airborne, cables or ground) featuring adaptations required to make the project economically viable for off-shore installation, aeronautical safety, environmental criteria for demonstration site selection and the definition of operational limits of the currently devised system. The WP also included definition of the cable swivels needed to avoid “in air” significant cable twist. WP 70 was dedicated to all testing (weather individual component testing) and the final proof-of-concept (i.e. on site, full assembled prototype testing) which ended up being performed at Ota Air Force Base, in Portugal, roughly 45km NE of Lisbon. WP 80 was dedicated to economic feasibility analysis, comparing an improved and scaled-up HAWE system (evolved from the prototype tested in WP70) with current state-of-the-art wind energy turbines to assess its economic competitiveness.
The HAWE consortium was arranged around the division of tasks presented in Figure 2-3:
Project Results:
1. Rationale
The main reason for embarking in a project like HAWE is that the energy return, from high altitude wind, must be sufficiently “rewarding” that the development of a completely new system makes sense. In latitudes similar to those of Europe (i.e. 36 to 58º of latitude North) the available wind energy at, for instance 2km, is by at least a factor of five times higher than that of typical wind turbine operating heights.
Wind potential grows with altitude, because wind speed typically increases with altitude, at 10km, for example, there is approximately 15 times more energy density than near the ground, with a very steep wind potential growth between altitudes of 3 and 8km. It is also of importance to understand what the characteristics of this wind availability are. As has been seen above the available energy contained in wind depends essentially upon:
• Latitude in the globe - wind is stronger between the 25º and 55º parallels, independently of the hemisphere chosen;
• Season of the year - wind is always, on average, stronger in the winter and weaker in the summer;
• Hemisphere - for comparable latitudes and seasons of the year the southern hemisphere displays stronger winds than the northern one.
Ken Caldeira and Cristina Archer’s research work shows that in the troposphere it is always advantageous to go higher in altitude even if also highly dependent of the latitude. For example, although at the same latitude, the northeastern coast of the US and the north of Italy, present different wind potentials for the wind speeds available in 95% of days (i.e the 95th percentile winds).
2. Concept description
The pursuit of high altitude wind energy has led to a number of different systems and configurations to harness energy from the greater power densities at altitudes higher than those used by conventional wind turbines. Since 2006, and with increasingly higher investment and technology readiness level, a system based on a hybrid (deriving lift from both aerostatic and aerodynamic forces) craft, tethered to the ground by a cable, has been in development at Omnidea. The lighter-than-air (LTA) structure takes advantage of the high aerodynamic coefficient of lift made possible through the Magnus effect, hence making this module a hybrid platform which we call the airborne module (ABM). The Magnus effect aerodynamic lift is generated by spinning a cylinder, allowing the system to take advantage of much greater values of aerodynamic lift and drag coefficients, when compared to conventional airfoil theory.
The system accesses wind resources at higher altitudes, by transferring the resultant forces of this wind power, through a multi-functional tether cable (structural cable including within it the functions of data transmission, electricity delivery to the ABM and buoyancy gas replenishment), to a ground station on the surface. The traction power of the ABM is used to produce electrical power through a pumping cycle, executed essentially in a 2D vertical plane in four phases:
1 – A power production phase (rising phase) in which the ABM is put into its maximum spinning velocity until the predefined operating altitude is reached; during this phase the cable is pulled and unwound from a winch drum coupled to a generator and therefore produces electrical power.
2 - A transition phase in which the rotation of the ABM is stopped as the maximum operating altitude is reached, resulting in a reduction of aerodynamic lift to zero, thus reducing the amount of energy required for recovery.
3 - An energy spending phase (recovery or winding phase), in which an electrical motor at the ground station rewinds the cable to its original position, bringing the ABM down to the starting position of the cycle. During this phase only the aerodynamic drag and the aerostatic lift of the ABM have to be overcome, but energy is still required to achieve it.
4 - As the minimum altitude for the cycle is reached, another transition phase takes place, with the ABM being accelerated into a spinning movement to begin a new power production phase.
The HAWE concept should thus be described as an airborne wind energy generator, capable of extracting wind energy from high altitude winds, through usage of a tethered, hybrid, airborne platform operating in a pumping cycle arrangement.
3. Component development
HAWE is mainly made of 3 separate (but interconnected) sub-systems:
• Airborne Module (“ABM”), comprising the whole flying platform, including the LTA balloon and envelope surface, the electrical hardware required to spin the LTA platform including motors and drivers, plus the hardware required to control the platform and communicate with the other systems (usually known as control box)
• The tether cable, including the functionalities of gas transfer from ground to ABM, electricity transfer media (i.e. electrical conductors), optical fibre for data transfer and the structural strength member itself
• The Ground station, which includes a winch, a sheave, a control hardware box with Graphical User to Machine Interface, plus the energy production and storage components (e.g. electrical generator plus a set of batteries, super capacitors or a flywheel)
The HAWE design is based around the following fundamental principles of operation:
• all individual components must be transportable in standard cargo containers,
• the concept must allow for modular (stacking) power increases.
• expectable lifetime should aim for 20 years
• the operating altitude must be changeable, at any time,
• low cut-in speed (2m/s) and high cut out speed (20m/s);
• The system will not operate under thunder storms, in which case it is reeled down and idled.
4. The Airborne Platform
The High Altitude Wind Energy (HAWE) production device is a machine designed to extract power from the wind harnessing available power at much higher altitudes than conventional wind turbines reach, or are likely to reach in the next 5-10 years. The task of designing a lighter-than-air platform that can operate at altitudes ranging from 200m to 2Km, with these altitudes changing every few minutes and with continuous 24h/day operation is an enormous challenge in itself. Thus several iterations were reqired to arrive at a stable and feasible prototype for testing in altitudes of up to 600m.
The main characteristics of the HAWE system are:
• High aerodynamic coefficients, made possible by the Magnus effect, which allow a high loading of the ABM even at low winds speeds of 2-6m/s without crosswind operation. To achieve these coefficients, the cylinder is spun faster than the incoming wind using an electric motor in the ABM.
• Lighter-than-air (LTA) design, enabling the ABM to remain airborne under conditions of no wind and stable under most wind conditions.
The study part regarding the ABM centered, mainly, on the following 4 areas:
• Dimensions trade-off for the ABM envelope geometry
• Aerodynamic study for the coefficients of lift, drag and moment calculation
• With all of the above, initial cycle calculations were performed
• Materials were also studied for the ABM
The altitude operating range mandates a certain pressure and temperature gradient between the interior of the ABM envelope and the atmosphere; it is thus the definition of the operating altitude range which determines the required strength of the ABM envelope, all layers included. The calculations of the internal buoyancy requirements are also not easy since they are dependent on multiple factors. Most of these calculations have variables which depend upon the definition of the operative requirements (cycle max. height and max. length, max. sustainable wind speed, etc…) but others must be estimated before the cycle begins (e.g. safety factor, etc…).
Due to the complexity of the calculations the team decided by a mix approach; a sum up of the expectable sub-system weight was performed, together with the expectable envelope and cable weights. After this was determined, specific dimensions were settled upon so that construction could begin; whenever new data would change the original project specifications, then the operational requirements would suffer from these changes. Furthermore, the system always requires a minimum relative pressure above 1.5KPa (i.e. 1500 Pa of overpressure vs. atmospheric pressure) in order to maintain shape and spin without major problems, therefore avoiding that its shape, when rotating, creates a “banana-like effect” around its center section, which induces multiple problems during the rotation.
Regarding volume-to-surface considerations, the ABM mass, for a given volume, varies with the cylinder’s length according to a relationship such as (1/l), while the weight of the ABM goes upwards as the l/d ratio is increased, as expectable by the variation according to (l/d). In summary a balloon with a high volume-to-surface relationship is desirable both to increase initial buoyancy (i.e. lower ABM mass for a given volume) as well as to maintain available buoyancy (i.e. reduce buoyancy gas losses). Still, in order to decide the diameter-to-length ratio of the test prototype, further studies were needed also, because it is the intent of this project to understand if the solutions being employed can be scaled for much larger ABM’s; hence several “typical” L/D ratios were investigated, such as 6, 8 and 10 (4 was deemed too small); for an L/D ratio of 6 full calculations were performed.
An extremely complex subject, regarding the ABM, is how to solve the huge forces, acting upon the ABM to gain volume as it goes up in altitude, towards a region of lower atmospheric pressure; this pressure gradient creates huge forces on the envelope, and must be countered by a Dyneema net, installed as the final 3rd layer of the ABM envelope. Nevertehless, due to the huge dimensions of the ABM, even Dyneema will reach its limits quickly, and hence a 1500-2000m maximum ceiling should be established. The problem increases proportionally with altitude, but becomes exponential with increased diameter; this is a further reason to keep a relatively low diameter, which in turns increases length for a given internal volume (i.e. buoyancy) requirement. Finally, there is the additional factor that small L/D will imply a high induced Drag which is something to be avoided as it reduces aerodynamic performance.
The materials selection is of paramount importance to the HAWE project as the ABM amounts for significant part of the cost and is the most likely to require frequent maintenance (due to exposure to an external environment). From earlier studies at Omnidea, the envisaged structure for the ABM should comprise 3 layers:
• a gas barrier layer or inner layer, where diffusion properties are most important,
• an outer layer, encompassing the inner layer, to withstand the pressure difference with the outside atmosphere without tearing, and to provide UV resistance,
• a structural net (a rope mesh) providing the bulk of the resistance and which is able to distribute the loads more effectively as, being a mesh, it can be oriented in a preferential direction
Candidate materials for the inner layer are thin film polymer combinations, basically arranged around a gas barrier plus bonding-friendly polymer arrangement (e.g. EVOH, a typical barrier film, being in a “sandwich” of 2 Polyethylene layers). Conclusions from the permeation tests induced the team to select a hybrid diffusion barrier material (PE/EVAL/PE) for the inner layer due to the low diffusion rates promised by EVAL, coupled with acceptable mechanical properties and good the bonding possibilities offered by the PE layers.
The outer layer materials are all of the polymer cloth type and, to join them together, one needs to stitch them. How easy it is to perform this task and how well both cloths are connected through this stitching, is what is mentioned as Ease of Work for the outer cloths.
A Dyneema structural net, with extremely high strength, was considered the perfect candidate for the outer net.
5. The Cable - requirements
The most fundamental input for the conceptual design of HAWE’s main cable is the correct perception of the amount of cycles required by HAWE’s operation. Each operating cycle is dependent from multiple variables, chief amongst which is the altitude interval of the cycle. In order to further proceed with calculations, a cycle altitude interval must be assumed as well as upward and downward speeds. But, with current knowledge it is fair to admit that, even for cycles, with altitude intervals as wide apart as 500 to 3000m (2500m of altitude difference) the upwards speeds should be assumed between 2-3m/s and downward speeds around 1,5x that, (i.e. 3-4,5m/s). Assuming an average of 2,5m/s of speed upwards and at a minimum of 60º of inclination, the “pure vertical speed” is around 2m/s. This means that 1250s are needed to go up in the cycle and 1250/1,5 s will be needed to come down, for a combined sum around 2000s. Hence, if a day has 86400s the minimum cycles which HAWE will perform in a single day is approximately 43 cycles. Likely HAWE will perform more, with an absolute minimum around 100 cycles per day more likely. But even taking 50 cycles a day this means over 17.000 cycles per year at 95% availability. It must be remembered here that HAWE’s specifications call for 20 years of operational life time.
To understand the problem further it must be estimated the amount of bending cycles, on the cable, per each HAWE cycle. This amount depends on the current layout of the apparatus and the driving winch. With the current layout we have 4 rope bending cycles, per each HAWE cycle, corresponding to bending on the storage drum and on the pivoting sheave, both on the upwards and downwards phase. This means 4 bending cycles per each power producing cycle, for a total of approximately 70.000 cable cycles per year. At 20 years of operational lifetime, this translates into a minimum of 1.400.000 cycles.
6. The cable - Structural part
There are several aspects to consider regarding the tensile properties of a fiber. Besides the importance of the tensile strength, it is also important to consider the fiber elastic elongation (related with the elongation at break). This last property is a measurement of the work (energy) necessary to deform the fiber.
Regarding cyclic bending-over-sheave (BOS) tests made for 3 material (steel wire, Dyneema and aramid) cables it can be seen that the Dyneema presents a linear behavior with a superior performance, when compared to the other materials under the analysis, so suitability of this material for loads of this typology is evident.
Initially Zylon appears as a good candidate but its strength is significantly impaired by exposure to high humidity/high temperature, and UV/visible light (precisely the operating conditions of HAWE) leaving Dyneema as a very good candidate in terms of abrasion, creep, UV performance and flex/fold resistance.
For the exterior abrasion protection jacket, the polyester fiber is the most suitable material.
7. The Cable- packaging
Traditionally, most cable designs place the conductor package in the cable center and apply reinforcement, around it. It is also common to add to the cable with a protective cover or jacket. In this layered construction the electric element is well protected from foreign object damage. A distinct approach is to place the strength member in the center (core) of the cable and arrange the conductor package around it. This construction has the disadvantage of placing the conductors on a less protected layer of the rope.
When the cable is put under load, the conducting core, the strength member and the jacket have to respond to the imposed tension. However each of these components has a very different stretch response. If the load imposed to the cable results into a higher extension than the said 0.45% elastic limit in the copper elements, the material will suffer permanent plastic deformation. After that plastic deformation, if the load is released, the material will only recover the elastic elongation, the excess conductor length has to be accommodated by the cable structure. If the cable structures acts like a compact assembly, the excess of copper length is forced to squeeze itself into a geometry that absorbs the plastic deformation length excess. The copper wire will form, at random locations, small double reversed bends. This behavior occurs each time the cable is again loaded, and will lead to ultimately failure, due to local fatigue effects.
For cables where the conductive copper elements are completely aligned with the strength member, only when the applied loads are very light, the copper can survive the loading cycles without plastic and fatigue issues. The conductivity of the copper wires is also influenced by the elongation of the wires. If the elastic limit is exceeded, the conductivity is altered and a more resistive area is created in the wire, which could lead to failure because of overheating and also damage the fiber strength member.
During the product development stage, several multifunctional cables concepts were developed, in order to integrate all of the required functionalities. For the development of this multifunctional cable, the low weight, diameter and high tensile strength were considered. Previous consortium member experience in the manufacture of high performance cables, indicates 12-strand braided ropes to the most suitable construction for the intended purpose.
For a HAWE system it is required a multi-function tether cable featuring a multi-layered cable with the following functionalities:
• A Structural “section”, for which breaking strengths, specific weight and fatigue performance are the crucial performance drivers;
• Power supply, to the Airborne Module, through copper electrical conductors running through the tether cable;
• Buoyancy gas exchange, between an high pressure He/H2 reservoir at the ground station and the ABM’s envelope
• data transfer, through optical fiber, for redundancy in communications
• The cable must also be possible to illuminate or be self illuminating
8. Ground Station - Winch discussions
From the multiple challenges that the HAWE concept poses in its development process, one of the greatest is perhaps the extreme operating conditions which the cycle execution, according to optimum aerodynamic parameters, dictates regarding the unwinding and rewinding speeds at the ground station, namely at the winch.
The winch system (that integrates also the control of all system setup) is the most important equipment at the ground station and, from an innovation point of view, one of the most relevant in HAWE. Per virtue of the magnitude of challenges considered above, the issue of developing an innovative winch was contemplated, at proposal stage, through an “innovation centered” approach in an initial delivered report where a trade-off, to reach winch optimum parameters (in terms of arrangement) is performed. The objective, through a component study interaction, is to find the winch arrangement which optimizes performance and is:
1. sufficiently rugged for everywhere, “always on”, operation for 20 years;
2. compact enough for transport without insurmountable logistic issues;
3. cost efficient enough not to compromise the overall project goal: undercutting current wind turbine energy production by up to 50%;
In a Winch the drum is the main component, and usually incorporates a power system interface (e.g. electric or hydraulic motor or generator/pump). The drum main dimensions, namely the diameter and its length are defined by the drum functions. The required drum diameter is calculated from the rope force and the power system torque, or from the rope linear and drum angular velocities.
Here, it should be noted that an additional restriction on the drum diameter is enforced by the rope durability. Namely, in order to minimize the rope damage caused by bending/straightening cycles, the drum diameter needs to be above a certain value, which is determined by the rope diameter and the rope material stiffness. Bellow the boundary value, the rope life expectancy diminishes significantly; hence from the rope lifespan point of view, the drum diameter should be as large as possible.
Once the drum diameter is determined, the drum length required to store the rope can be calculated. The number of rope layers is usually limited to 4 or 5 as a compromise solution, between the two opposite requirements (i.e lower length and longer life time), must be found.
In order to spool the rope onto the drum in a correct manner, i.e. without crossing the rope over itself or leaving empty spaces, the attack angle of the rope spooling (with respect to plane perpendicular to the drum main axis) must be kept at low values. Depending on the dimensions of the drum, the maximum allowable spooling angle requirement can be achieved without any additional special intervention. Due to the more complex motion of HAWE's ABM and a high probability of using multilayered drum (due to the long rope required), a guided spooling is necessary. The simplest solution to the spooling problem is to set up a separate pivoting sheave, placed at a given distance from the drum.
For HAWE’s constraints the sheave needs to be positioned at least 20 “drum lengths” away from the center of the drum. As an alternative, the pivoting sheave can be connected to the winch by means of a console, thus being rigidly connected to the winch. Single layered drums require no additional spooling guidance; however, multilayered drums usually require additional spooling mechanism to enable correct spooling of the rope in each layer.
An additional requirement for the winch intended for HAWE application is related to the special tether cable, which is to encompass power, signal wiring and a gas tube for ABM replenishment. Consequently, in order to connect the wiring and tube from the rope to the ground station systems, a special slip ring system needs to be used. Depending on the final system solution, at least one drum side has to be unoccupied by full shafts to place the slip ring system. Solutions using a hollow shaft with sufficient inner diameter may have certain advantage over the full shaft solutions, because the slip ring system can then be placed on the motor/generator side (far away from the winch), thus leaving one of the winch sides free for any other purpose (e.g. for double generator connection, or for using separate slip rings for power/signal and gas). Three different configurations were investigated by Rapp Hydema for the system drive choice:
1. The mechanical driven flywheel concept alternative; with good potential for future serial produced HAWE power generation plants. The cost impact and detailed adaptation of technology in such a system is unknown at this stage.
2. Several HAWE power generation plants working “out of phase” - this could present an alternative without energy storage requirements. To obtain a constant power generation to the net/grid, at all times, two (or more) systems must be in a power production phase per each system in the recovery phase.
For the proof-of-concept trials at Ota a direct pull winch, in a system without energy storage, was used employed, based on the following advantages:
• Will provide constant torque and consequently variable pull and speed versus the pulling diameter;
• If the drum is long, the difference in pull/speed on first versus top layer will be smaller;
• If the distance from the center of the drum to the fixed block is 20 times the drum length there would be no need for a level wind system (self-spooling);
• Will get 1 cycle (360⁰) at full load – not considering the level wind system;
• Limited mechanical contact between the rope and the drum for a winch without level wind;
• Physical contact between each wrap of the umbilical when spooling on to drum;
• Less complex design (in comparison with a traction winch);
The arrangement selected was a pivoting sheave as a separate unit plus a level winder. The HAWE system used a PTS Pentagon Research hardware control equipment.
9. Ground Station - control software
The ground station control is intimately related to the cyclic nature of the system as well as to the issue of energy storage (to "even out" the energy output to the grid, stemming from the pumping cycle nature of power production). To this end a report, including a comparative review and assessment of mid-size energy storage systems (< 10MW) was performed comparing flywheels, hydropneumatic accumulators, electrochemical batteries, and ultracapacitors, for criteria such as efficiency, costs, capacity and reliability. The study identifies ultracapacitor-based energy storage systems as the most viable for self-standing HAWE energy storage solutions, excelling in terms of running costs, acquisition costs, availability, efficiency, bulkiness and durability (i.e. system longevity) at low-to-mid altitude ranges of the ABM (i.e. when a relatively large number of charging/discharging cycles is needed). Further conclusions include the flywheel energy storage being competitive in the low-to-medium altitude range and sodium-sulfur and flow batteries becoming viable solutions whenever the number of cycles is significantly low.
With the modeling allowing calculation of the 2D cycle trajectory, multiple electromechanical configurations and energy storage systems, for the ground station, were simulated to achieve the optimum operating parameters regarding winding and rewinding speed as well as spooling speeds and torque.
The high-level controller manages the power flow to the energy storage subsystem, in order to provide a constant grid power flow over the whole HAWE cycle, while satisfying the condition on storage state-of-charge (SoC) sustainability. The high-level controller has been first developed for the case of electrical storage system, where the battery, ultra-capacitor and flywheel storage technologies have been considered. The controller includes feedback and feed forward actions, and a low-bandwidth filter used in the feedback path to average the SoC over the HAWE cycle.
10. Theoretical results, simulations and calculations
Since no calculated aerodynamic coefficients were available initially, to perform basic cycle performance calculations, Reid’s (1925) set of aerodynamic data for cylinders subject to Magnus effect, were used. The performed aerodynamic simulations provided coefficients for the cycle simulations and optimizations, thus giving insight on power production and recovery phase of a single module. Single cylinder was simulated, as well as two cylinders with varying distance between them.
The impact of turbulence closure on results is very large and requires estimation, which was achieved employing LES (Large Eddy Simulation) using the Coherent Structure Model (LES-CSM) approached. Results of aerodynamic coefficients for single smooth cylinders show a larger than expected amount of lift force, obtainable from cylinder rotation. Results for two cylinders indicated that extra lift can be obtained, but that small distances between the cylinders (less than 2-2,5 diameters) should be avoided due to a significant impact on the aerodynamic coefficients for each individual cylinder.
A “realistic” model, developed by Omnidea and DTU, accounting for factors such as:
a) the basic 2 Degrees of Freedom (2DoF) ABM model including 3 rope sub-models of different complexity,
b) formulation and implementation of HAWE system control variable optimization problem including preliminary results and insights,
c) the multi-body model concept intended for ABM passive stability analyses and more rigorous simulation studies including control system robustness analysis, and
d) an analysis of multi-balloon system-based control authority over yaw (and roll) dynamics.
was developed and further evolved, for a final iteration with the university of Zagreb. At its final stage it allows representation of a full cycle in 3D representation, as well as much more advanced cycle performance calculations. A real-time forecast system, capable of providing real wind conditions over multiple locations (proposed as potential HAWE testing sites), was implemented. Data was subsequently treated, through DTU’s forecasting software (WRF) to provide simultaneous measurements of time, height, free stream wind velocity, wind direction, temperature and air density and viscosity. This WRF modeling was implemented with Omnidea and has been updated to include several real wind speeds measurements at Ota test site.
To achieve a pre-commercial unit design of 80 kW of average power (roughly 200kW of peak power), based on the same simple concept of one ABM coupled to one ground generator, four rotating cylinders will be stacked in a single ABM. As the operational requirements mandate that the cable speed for the recovery phase is limited to 6m/s, this limits the maximum Cp that can be achieved to 1.235 (vs. 1.603 for a higher recovery speed). For simplicity of control and actuation, the cable speeds and the angular speed of the cylinder are constants.
The simulated cycle achieves peak power, at the end of the production phase, because that is the region with the highest wind speed. Therefore the average power of the generation phase will be lower than the peak power. For the cycle simulated, the peak traction power is 203kW and the average is 166kW. The total time for the generation phase is 568s. The average power spent to rotate the cylinder is 6kW, mostly due to inertia, whilst the power spent during the recovery phase, which lasts 398s, is 38kW. The average traction power for the whole cycle of 16 minutes is therefore about 80kW.
The energy produced in this cycle is over 5 times higher than the whole energy consumed (recovery plus rotation) and over 6 times (if recovery only is counted). It is expected that, with improved control strategies this value could be increased to 8. These simulations show that the components in the system need to be oversized to accommodate the fact that the peak power is higher than the average power by a factor between 2 and 3 (the factor for the actual cycle simulated is 2.52).
According to initial simulations performed at Omnidea for a single HAWE system, the ratio between the system’s maximum power production, achieved during the upwards phase of the power cycle, and the average power production per cycle (taking into account the transition periods, the energy spent to recover the ABM to the original position and the energy to rotate the ABM) can be up to 2,5; i.e. average power production is roughly 40% of the cycle’s peak power. This ratio influences the choice of components for the ground station and ABM, depending on the configuration of the power generating devices. Also, because the electrical grid requires energy to be fed in the most constant way possible, it is always desirable to “flatten” the energy feed-in curve; this curve is not flat at all for a single HAWE system, due to the pumping cycle’s nature. This then creates 2 issues:
i) If a single HAWE system is operated, then an energy storage solution is required to perform the ABM recovery, or a connection to the grid must exist;
ii) In the example previously presented, despite the production of “only” 80kW on average, some components must be dimensioned to accept over 203kW peak power and this is without considering factors of safety on the equipments.
Hence, compared to the basic configuration with a single HAWE system, alternatives (employing two ABM units) were analyzed:
a) wind farm configurations, by having multiple HAWE systems in a single location, with the main advantage being a reduction (or elimination) of the need for energy storage.
b) dual-winch with a single-generator configuration: this configuration allows installation of a generator unit more closely matched to the average produced power, significantly reducing the investment costs per unit of injected energy into the grid, compared to a single HAWE system configurations. The results can be better understood from Deliverable 50.3.
Apart from the basic, single-unit HAWE system configuration equipped with electric storage system, two more advanced configurations have been assessed for systems containing two ABM’s and two winches: dual generator configurations and single-generator configurations.
However, the single-generator system under a dual winch arrangement is only semi-controllable since there is little possibility of independently performing speed control for both winches; this leads to situations where the imposed winch speeds lead to non-optimal load-speed combinations.
Hence a third alternative has been proposed: a single-unit configuration based on a compact transmission arrangement with the use of direct flywheel (or hydro-pneumatic storage solution) for energy storage. This system provides a good blend of favorable overall performance characteristics with simple but enhanced controllability and reasonable investment costs. Similar structure of the high-level controller has been applied in the case of ground station configuration based on a compact transmission, including a direct flywheel storage connected to the winch through the gearbox. Here, the winch speed cannot be directly controlled during the ascending or descending phase (there is no low-level controller), because the winch speed is determined by the approximately constant flywheel speed. This can affect the ABM control performance, particularly for optimal performance control which happens when the ABM achieves near-vertical motion (meaning that there is almost zero drift with wind and thus no loss of energy due to decrease in relative wind speed). On the other hand, the transmission clutches are properly controlled (shifted) to provide smooth ABM reversing, with non-excessive clutch energy losses. In the case of planetary gear-based compact transmission, the winch speed can be effectively controlled by means of a low-level generator speed controller. This is because the planetary gear cinematically decouples the winch from the flywheel. The high-level controller has a more complex structure due to the more complex kinematic arrangement of the transmission.
11. Proof-of-concept and on-field results – test site preparations
Tests were conducted at Ota Air Force Base to be performed on days with the wind coming mainly from a “North oriented direction”, due to the specific needs to launch HAWE. The restrictions to operate under the predominant wind directions means that the entire length of the runway, (up to 2500m) is available as a safety margin in case of system failure, thus ensuring that whatever incidents occur with the prototype, it will always be confined to military facilities.
The assembly of the winch at Ota required some terrain preparation work such as ground leveling and construction of two concrete slabs. The main concern was to ensure proper alignment between the winch and the sheave. To satisfy the demands for electricity during operations at the field site a diesel generator was rented.
The HAWE concept operates at different altitudes during a single production cycle. It is therefore important to have reliable measurements of the wind conditions at the different heights of the Airborne Module. The wind data at Ota was provided by LIDAR providing wind profiles up to 200m and an ultrasonic 2D unit fixed at a 10m reference height. One of the objectives of having the two independent units to gather wind data is to be able to validate the forecast models developed by DTU and to assess the correlation between the 10m reference and the wind profile with altitude, so that reliable data on the wind profiles, without having the need for the expensive LIDAR unit, can still be obtained.
Prior to any regular tests or outdoor operations, a check list sequence was created to check proper functioning of all systems avoiding situations in which tests would be carried out without being able to, for example, record or log the obtained data. To assist these tests, validation of electromechanical components, gas-tightness of the ABM and confirmation of the communication capability between the ground station and the ABM was necessary. Also, the test site needed to be supplied with adequate mooring structures to assist the anchoring of the ABM. Additionally these structures could be used not only during pre-tests but also to support the phases of outdoor tests, in particular assistance in the takeoff and landing phases. Three support structures are required for operation:
• ABM trolleys to move it from inside the storage tent to the tests area in safety.
• Two small winches to allow for three mooring points for the ABM, before takeoff or after landing.
• Two metal framework supports to accommodate auxiliary cables from the small winches that constitute part of the mooring system of the ABM.
12. Proof-of-concept and on-field results – wind study at Ota
In order to have an idea on the importance of atmospheric parameters, besides the wind sonic station also a LiDAR equipment as well as the WRF (Mesoscale Weather Forecast Model) were used, both supplied by DTU-WE, to study wind at Ota. The LiDAR performs actual measurements, once every 18 seconds, for varying altitude levels until 200m while the WRF is only a forecasting tool, giving a forecast value once each hour and for a very broad swath area (nearly 1,5km2). The WRF can provide estimates until 2000m for 14 different height steps.
Low winds (< 5 m/s) occur during 55% of the time at an altitude of 10 meters and approximately the same at 15m; it is when we go up, towards 100m, that things change as this percentage drops to around 25% only. Between 100 and 200m the trend is for winds stronger than 10m/s to become much more probable; according to the WRF model, from 200 to 300m this trend is completed in the Ota test site as, all the way to 1000m, only winds in the 15 to 20m/s range clearly gain in probability. According to data provided by the WRF model for the Ota site, for altitudes between 500 and 1000 m, the WRF model predicts a significant increase in wind intensity, but also a shift in the importance of wind with the SW-W direction becoming more and more important as we near 2000m of altitude. This comes mostly from a decrease in the NW-N direction, while the W-NW starts strong, loses importance between 300 and 1000m and returns to its initial magnitude between 1300 and 1600m.
Analyzing the data, we see that, as we go up in altitude, the highest wind frequency peak moves to the right (i.e. towards stronger winds) as we go up in altitude. This can be considered the "normal" or "typical" curve format.
The corresponding cumulative annual wind energy from LiDAR measurements, shows that the energy available at 10m of altitude, 104m of altitude and 199m of altitude is respectively 0.5 2.5 and 4.7 MWh per m2 per year. Another way of interpreting these is to say that, at 10m of altitude, nearly 36% of all wind energy comes from winds below 5m/s, a percentage which decreases to 7% for 104m of altitude and even further (to 4%) as we approach 200m. Similar interpretation reach over 90% of wind energy from winds below 10m/s at 10m of altitude, over 80% at 100m and slightly over 50% at 200m. Above 200m there is no LiDAR data and, hence, we must rely on WRF data.
Thus, in order to get a better perspective on the agreement between both data sources, a comparison, at the highest altitude level available from both sources (200m) was made. The comparison reveals the issues with the WRF model: instead of having a total 8MWh/(m2.year)] predicted by the WRF model, the LiDAR measures no more than 4.7[MWh/(m2.year)] i.e.: only 60% of the value given by the WRF Model. We expect such a correction factor (60%) or even higher could have to be applied to the WRF data if a more conservative wind energy availability is to be computed for higher altitudes.
After having understood how the LiDAR and WRF data correlate, the histogram and cumulative histograms for the 300-1000m of altitude intervals reveals that as we go up in altitudes between 300 and 1000 meters the shape of the wind probability graphs does not vary significantly in this particular site (hence probably one of the main reasons why it was selected for take-off and landing of military aircraft), although until 500m the probability peak is still to the right of the probability peak at 300m. Nevertheless things are much less clear for 700 and 1000m where even "double peaks" can be seen, i.e. the probability function peaks twice, in both cases between 5 and 10 m/s.
This translates into the fact that, at least at Ota, although 600m is teh maximum reachable due to air-traffic regulations a system designed to operate, for example, up to 1Km above ground, would benefit more from the constant characteristics (smoothness) of the cycle, then from a clear increase in energy availability.
So how does the system compare with its theoretical maximum, in terms of energy? A simple calculation can be performed: applying the correction factor (between LiDAR and WRF) presented, and operating at 500m, HAWE could theoretically harvest around 8 MWh/(m2.year) (i.e. 13 MWh/(m2.year) x 0,6). This can be used to estimate that a prototype, such as the one tested with roughly 40m2, could harvest not more than 320MWh per year, or an equivalent energy production approximately 3kWh per each 5 minute cycle (i.e.: 36kW of power).
13. Proof-of-concept and on-field results - mechanical layout efficiency
The specifications for the ground station present in the field tests at Ota proving grounds were for an ABM with a maximum generated power during the production phase of around 60kW to achieve an average cycle power of 20kW. Nevertheless in testing with the current prototype the power figures were much smaller due to its much reduced dimensions, with a maximum of 10kW during the recovery (Ground Station electric machine working as a motor) and 3kW during the production phase (Ground Station electric machine working as a generator). Consequently and because there was clear evidence that the conversion of mechanical energy into electric energy, at the Ground Station, was not being done efficiently, a no-load test was performed.
For the no-load test of the Ground Station, the cable was completely reeled-in onto the winch drum and then the drum was rotated. Since the cable was not connected to any mechanical load, the only power consumption under constant speed was due to power losses of different components associated with the Ground Station system, namely on the winch gears, the driver and the resistor used to dissipate the energy produced, since the whole system is not connected to the grid. During the no-load tests the winch showed a power consumption of 16.4kW at 6m/s and 18.9kW at -6m/s (reverse direction). These power losses represent a permanent base load power consumption of 23% of the maximum power (60kW) just to spin the drum. Although this value is already quite high, an even greater problem occurs when operating far from the maximum power regime as, in terms of percentage, the power losses becomes much, much higher compromising any and all small net electric power production.
At this stage it was clear that there was a problem with one of the components subsystem, compromising the experimental potential proof of concept. Analyzing the data to analyse the efficiency of the Ground Station system, the Mechanical power and system efficiency can be calculated from the equation which states that the mechanical power equals the multiplication of the force (at the load cell in the end of the cable) by the cable speed.
The regime of operation of the Ground Station in the Recovery phase (electric machine as motor) spans approximately from 0 to 10kW with the efficiency achieving a peak of 50% around 9kW of mechanical input with a plateau of around 30% from 2kW to 8kW.
The regime of operation of the Ground Station in the Production phase (electric machine as generator) spans approximately from 0kW to 3kW. The problem is that, most of the time, the mechanical power input is not even enough to overcome the Ground Station power losses, so the installed winch would actually be working as a motor (i.e.: spending energy) when it should be working as a generator.
In the Recovery Phase (i.e.: system operating as a motor) for a given cable speed there is an increase in the efficiency with an increase of the mechanical power input as expected. Since power = force x linear speed, and there is a direct correlation between speed and voltage as well as between force and current, it can be concluded that the installed winch system will have higher efficiencies for higher cable forces almost regardless of the cable speed. The same analysis seems to be true for the Production phase. To support this conclusion it is important to understand that, for both phases, the efficiency is almost linear with the cable force and that the efficiency in the Recovery phase has a peak, at 80%, for a cable force around 4500N. Because, during the production phase the force required to overcome all the losses in the Ground Station is 1.5KN (i.e. only above this value of force is the installed Ground Station actually capable of producing electricity) then we can understand the reason behind the relatively small amount of electric energy produced: In the installed winch for testing, the efficiency is higher for low cable speeds and high cable forces, which is the exact opposite of what was needed in HAWE.
Unfortunately the threshold of 1500N (roughly 12.5% of the maximum load) to start producing electricity is too high to be overcome in almost any day of testing. Hence it became clear the components of HAWE’s ground station were not adequate for the typical operating regime of the prototype; more so, during the final season of testing the lack of ground station “suitability” and “efficiency” was the single largest contributor to not having been possible to execute cycles with net positive electrical energy generated. Nevertheless because the prototype was fully instrumented it was possible to achieve proof-of-concept by proving that there were cycles executed producing net positive mechanical energy. This net energy production is demonstrated by extracting the forces and speeds exerted on the tether cable (i.e.: mechanical power).
14. Proof-of-concept and on-field results - analysis results of a representative cycle
Some cycles were performed between the 23rd and the 24th of January 2014 and, cycle number 4 is the cycle which will be analyzed in detail:
-The duration of the cycle was of 70 seconds;
- the average cycle height was of 191m;
- the wind at average ABM altitude was of 6,23m/s;
In this case the mechanical power produced at the winch is around 1kW, despite a 0,35kW power consumption to rotate the ABM.
Cycle number 4 presented a rather high mechanical power, and the 2nd lowest wind at ABM height when compared to other cycles. Cycle altitude can be considered average as can the ABM power consumption when in comparison to other cycles. In this cycle, the cycle did not begin and end exactly at the same height; the cycle is considered performed when the cable length, at the end position, is the same as the initial cable length. Minimum cycle height was around 150m while maximum cycle height reached around 220m. The 3D trajectory plot shows a drift with wind of about 50m for an altitude difference of around 70m.
Cylinder rotation was kept constant at 50 rpm during the whole power production phase, being reduced to 10rpm during recovery. This means that recovery was not performed with a non-spinning ABM, but rather with a still spinning ABM, which increased the power needed for the recovery phase and, consequently, decreased overall cycle efficiency significantly. This is a necessity in HAWE in order to keep flight stability (i.e. otherwise recovery becomes very difficult and takes much longer with the possibility of crashes against the ground). Nevertheless, the main objective of the tests was to validate the simulation models and running the latter with similar parameters showed the same results therefore demonstrating that, under the proper parameters, HAWE produces net positive energy.
Cable recovery speed was around 2,2 m/s with an upwards cable velocity around 1,5m/s during unwinding (power production phase). Hence the “rule of thumb” of the velocity being 1,5x greater during recovery (vs. power production phase) was approximately achieved in this cycle, although both speeds were below that considered optimal. Power needs during recovery were around 1kW, however, at the beginning of the recovery phase there is an important power consumption with a peak of approximately 4kW. This peak significantly worsened the results and was due to the fact that, when recovery was started, the ABM was still rotating at nearly 25-35rpm due to significant inertia to slow the ABM rotation. Nevertheless, as previously mentioned, the results validated the theoretical model and the computational simulations.
The results obtained confirm the validity of Flettner’s work in spinning cylinder’s, dating from almost a century ago (1925). As expected the lift coefficient obtained was slightly below the theoretical values, with the same happening on the drag coefficient side. The measured CL/CD curve has an evolution which greatly resembles literature values with two minor exceptions. One of these being that the measured drag coefficient suddenly shows an erratic decrease and the other (towards the final phase of the cycle) the measured lift falls lower than expected. These two cases of measured lift being below “expectable” can most likely be traced to the misalignment of the ABM with incoming wind and by the system’s “elasticity”, measured by the cable catenary, which causes delays and fluctuations in the system force, both in intensity and in direction.
Throughout most of the upwards phase the operation is at a speed ratio, between 1 and 1,5. During recovery, as the rotation is reduced severely, the speed ratio falls below 0,3. It must be reminded that the HAWE system was designed to operate with speed ratios close to 2,5-3m/s during power production phase, with an aerodynamic optimum between 3 and 4. Nevertheless operation at these speed ratios requires the ABM motors to be capable of sustaining an rpm level around 90rpm or higher. This (90rpm ABM rotation) was achieved during the final phase of testing, after October but, unfortunately, the LiDAR equipment was no longer available at Ota.
The system model is a 2D model, which takes into account that the balloon is always aligned with the wind direction. We know that during the upwards phase, the tension in the cable ensures this alignment but, during the descent phase, tension on the cable decreases significantly, making the trajectory unstable and extremely susceptible to lateral displacement (sideways) easily becoming misaligned with the wind direction.
The magnitudes of the different powers which influence the HAWE prototype during the analyzed cycle are:
• Average power generation, during the power production phase is 2,4kW;
• Average power spent during the recovery phase is 1,4kW;
• The average power needed to rotate the ABM was of 0,35kW, leading to an actual “net” cycle power of 1,05kW (excluding transition period).
(Note: Transition period is always constant and has a high influence on average power depending on the length of the cycle)
These test values were obtained at 5,4 m/s of average wind speed (this average wind is the true average wind throughout the whole cycle), which is clearly below the “nominal design wind conditions” of 8m/s (while standard wind turbines are designed for nominal operation of 10 to 12m/s of wind speed). Because extractable power from wind increases cubically, if everything else is held constant, a factor of above 3, in average power production could be achieved for constant winds around 8 m/s. With the ABM spinning at above 100rpm, this factor could be increased to above 6, leading to, potentially, an average power production around 10kW during the upwards phase.
One very important factor was also evident from testing this prototype: since HAWE is a wind power conversion system that consumes power during part of its operation, it is extremely important to ensure that specifications are met at all subsystems level; otherwise it may happen that one system is jeopardizing the entire system performance and its viability in achieving the objective for which it was designed and manufactured.
15. Proof-of-concept and on-field results – conclusions from testing
An average net mechanical power of 1kW was measured and the results were according to the computation simulations under the same parameters, thus validating the model, the mechanical power could not be translated into electric power due to the extreme inefficiency of the winch.
Despite some testing and grinding issues with the prototype, and all above described system limitations, it was possible to validate the models and therefore provide proof-of-concept with the prototype.
Some issues faced:
a) As explained in section 3.5.3 the ground station efficiency was very low, with the winch (in particular) requiring a very significant pulling force (upwards of 1.5kN) to be able to produce net positive electrical power. Due to normal design safe requirements this component was already over dimensioned by a factor of between 2 to 3. But, because the actual ABM prototype tested was roughly 1/4 of the initially intended size to be tested, the whole ground station ended up being grossly over dimensioned (very likely by a factor of over 10) and this meant that no net positive electrical power cycle was achieved. Calculations thus had to be performed for purely mechanical power considerations, from the load cell measurements taken at the tether cable and swivel point. The electromechanical concept at ground station must, nevertheless, be improved to allow the system to unleash its full potential as well as to optimize its operation parameters;
b) During the upwards (power production) phase, a maximum of only 55rpm could be reached, during the January-February 2014 test campaign. During the same year’s summer test campaign this restriction was overcome and testing with the ABM rotating at 80rpm has been achieved, increasing the Magnus effect speed ratio from 1,5 to around 3,5 (achieved with rotation at 80rpm).
c) The ABM prototype’s net buoyancy is below 50N, which means that, as soon as ABM rotation is stopped the ABM envelope will immediately start to fall, and do so in an unstable fashion compromising the envelope’s structural integrity as well as that of the cable and compromising also the safety of the personnel on the ground station operations; for this reason a rotation to the ABM was imprinted even during recovery, implying much greater effort (i.e. more power) needed to perform recovery;
d) The motors developed for the ABM prototype, by Omnidea, proved a good option with an excellent power-to-weight ratio and the capability to operate above 600V reducing losses through the cable; nevertheless, as this was a task that was later necessary to add, to speed up the development still during the HAWE project, the drivers were not yet designed to also incorporate the function of working as a generator. Therefore the motors could not be used to break the ABM’s rotation; consequently when the ABM was already being recovered at 2 m/s the balloon would sometimes still be rotating at above 30-35rpm, leading to an initial 4kW peak power in the recovery phase. The semi-automatic control system operation during these tests, was not flexible enough to allow changing the synchronicity between balloon rotation and cable speed, minimizing the power consumption due to the balloon’s inertia;
e) Instead of the expected 5-10% energy consumption from the ABM (to keep it rotating), nearly 20% of the average power production was needed. This effect is due to the balloon bending between its “cantilever type” fit in both ends, an un-anticipated effect and is considered, by the HAWE team, a very important experimental result as it is now known that it has to be minimized. Bending can be minimized with higher internal pressure, with bigger rims (structural reinforcement of the tops) or a different rotation system that has been developed throughout this project and tested at sub-system level. In this alternative ABM rotation system, the cylinder is anchored in a third point (through a ring in the middle section) and rotated through this ring. Further info can be obtained in Deliverable 10.4.
16. Economics and financials of HAWE - Assumptions
As defined in previous sections, the HAWE system is constituted by an airborne module (ABM), a ground station a tether cable and a control system that includes communications. In addition, other components such as grid connection, substation and other auxiliary items are also part of the system. The cost structure considers already significant cost reduction when compared with the costs of the prototype, namely the cost of the winch. The main cost drivers are through industrialization of the solution and reduction, raw material and labor required to produce the equipment. It is important to mention that these costs do not reflect adequately the learning curve effect for producing several units. A summary of assumptions can then be provided:
• The main Operation and Maintenance (O&M) activities required are the replacement of the ABM envelope and tether cable, the envelope due to aging and the cable due to cycle fatigue. As a base scenario, it is considered that the cable and part of the envelope is replaced every 2 years. In addition, periodic preventive maintenance shall be carried every 6 months;
• The evaluated project base scenario consists of a “15 units HAWE farm”, each individual unit with up to 250 kW of capacity for a total of 3,75 MW. It was estimated that the farm would require the installation of an aerial electrical line with 5 km of length. It was also assumed that the devices will lose 5% of performance in the first 10 years of operation and 15% up to the 20th year of operation.
• The cost of each component was estimated based on the prototypes’ experience and quotes from supplier and considering that the technology is in a pre-commercial, resulting in a CAPEX for 250kW units of 150KEUR
• As a base scenario, it is considered that the cost of Operation and Maintenance (OPEX) of the farm would be of 5,3%/year of the total investment (CAPEX), which totals approximately 253 k€/year.
• In order to have a comparative analysis against wind turbines a 35% capacity factor was assumed (i.e.: 3.066 equivalent hours of operation at nominal power), which is understood to be very low and not taking advantage of the high altitude winds stability, likely leading to a very conservative approach.
From the economical and financial assumptions the following were based on analysis of the available technical information.
• The Feed-in-Tariff for this project was based in the current Portuguese legislation and as a base case, the tariff is 74 €/MWh.
• Other scenarios, considering experimental feed-in-tariffs, ranging between 162 €/MWh and 208 €/MWh were also modelled in the sensitivity analysis. In either case, the Feed-in-Tariff is limited to 33 GWh/MW or 15 years. After this period, the tariff is the pool price, which is assumed to have an average price of 50 €/MWh.
• In this case, the depreciation was calculated on a straight-line basis, meaning that the capital cost of the equipment was spread evenly over its depreciation period. The depreciation periods were determined according to the prescriptions of Portuguese legislation.
17. Economics and financials of HAWE - Results
With the base assumptions, the project proves to be viable, with a positive NPV and IRR higher than the minimum Return Rate required (WACC). The payback period is 8 years, which is acceptable. The Levelized Cost of Energy (LCOE) is 64,2 €/MWh, which is very competitive with alternative renewable solutions, in fact almost achieving grid parity.
Based on the analysis above, several scenarios were defined:
• Scenario 1 – Base scenario
• Scenario 2 – Increase in energy production by 10%
• Scenario 3 – Reduction in energy production by 10%
• Scenario 4 – Increase in CAPEX by 10%
• Scenario 5 – Reduction of O&M costs by increasing the life time of the cable to 5 years
• Scenario 6 – Reduction of O&M costs by increasing the life time of the cable to 10 years
• Scenario 7 – Increase of the feed-in-tariff to 162 €/MWh
• Scenario 8 – Increase of the feed-in-tariff to 185 €/MWh
• Scenario 9 – Increase of the feed-in-tariff to 208 €/MWh
• Scenario 10 – Reduction of insurance cost
The main results were:
• The HAWE technology is very sensitive to the energy produced. A change of 10% in energy production (scenario 2) allow us to increase the IRR by almost 300 basis points (BPS).
• a decrease of 10 % in energy production (scenario 3) reduces the IRR by 270 BPS, in this situation the project is not profitable, the NPV is negative and the IRR is lower than the minimum required return rate.
• The same conclusion applies if the CAPEX increases by 10% (scenario 4). The NPV of the project is negative and the IRR is lower than the minimum required rate of return.
• The HAWE technology is also very sensitive to the O&M costs (scenario 5 and 6). The base scenario O&M costs are high. If a reduction in O&M costs is achieved, it will have significant impacts in the financial indicators of the project. If a reduction of O&M costs from 5,3% of the total CAPEX to 2,3% is achieved the NPV more than doubles and the IRR increases 350 BPS. The payback time of this scenario is 6 years.
• The HAWE technology is also very sensitive to the feed-in-tariff (scenarios 7, 8 and 9). If the project can obtain an higher feed-in-tariff, the financial indicators improve significantly and the IRR of the scenarios are respectively 37,8%, 46,0% and 54,1%.
• The last scenario (#10), considers the reduction of the insurance premium. Such reduction will allow improving the financial indicators, but it is not as significant and the previous scenarios.
18. Concluding remarks
To conclude, it is very important to for the success of the technology, to implement it with the possibility of extracting the wind energy from high altitude, and therefore achieve high capacity factors which is the highest driving factor to render any project based on HAWE technology as economically attractive as possible. As an example, if there would be a deviation of additional 10% on the base scenario on both capacity factor and CAPEX, then in spite of the negative contribution of the latter (i.e. increase in CAPEX), the overall result would be an improvement in the economic attractiveness (as the increase in capacity factor outweighs the CAPEX effect), meaning that the project CAPEX has a much lower impact proportionally than the capacity factor. Decreasing costs of O&M also has a major impact in improving the return on investment by reducing the payback time; however O&M costs requires a higher technology readiness level (TRL) in order to be properly assessed.
The HAWE team will proceed with the R&D on this technology and with the TRL increase still needed, hopefully until a successful market implementation.
Potential Impact:
Current estimates of the power factor of the system developed in this project, derived from simulations performed and from preliminary results obtained in testing at Ota base, and present in EDPi’s economic feasibility assessment study, indicate that HAWE can be competitive with conventional wind turbines with the distinct advantage of being able to provide electricity on a continuous basis by operating at varying altitudes to capture high altitude wind energy. In the mid to long term, HAWE is expected to meet the sustained demand for energy worldwide as it can operate in a greater range of environments than conventional wind turbines and is easily deployed offshore. The successful development of the HAWE system addresses the global concern for renewable energy, contributing to the SET plan of the EC and furthering Europe´s position as a leader of innovative technology.
Since 2010, when the HAWE project began, the overall wind market has continued to grow, although the 25% Cumulative Annual Growth Rate of the 2000-2009 period could no longer be sustained. After new installations of wind power stabilized in the 2009-2011 period at around 40GW per year, 2012 saw an increase to 45GW followed by a decrease in 2013 to 35GW of new installations. This means that, in 2013, new installations were back to 2008 levels. In terms of worldwide accumulated installed wind power, in June 2014 roughly 336 GW were installed. In the period 2009 to 2013, wind energy finally overtook biomass as the 2nd largest source of renewable electricity and electricity from wind is fast approaching 1000TWh of worldwide electricity generation roughly one third of worldwide electricity generation from hydro. In the same period (2009 to 2013), new turbine installations in Europe have been steady at about 10-11GW per year and Europe grew, in terms of overall wind installations, from 76GW in 2009 to 117GW in 2013.
In 2010 the wind market represented 0,26% of Europe’s GDP with a turnover of around €32,5Bn from direct (€17,6Bn) and indirect (€14,9Bn) employment. Of this turnover, nearly €9Bn were exported, with imports of only €3,6Bn, i.e. in the wind market exports were approximately 2,5 times as large as imports! Employment stood at nearly 240.000 jobs (135.000 direct jobs plus 105.000 indirect jobs). For each €1Bn invested in the wind energy industry, 21.000 jobs were created per full year (source: EWEA). Furthermore, by dividing the amount of created value in the direct employment of turbine manufacturers, component producers and service providers, we get an average of 82.000€ per person (source: EWEA), per year in the wind market, making it standout as a value-added, export oriented market.
Nevertheless, although Europe is still the region of the world with the largest installed wind capacity, China has overtaken Europe in new wind installations and should overtake Europe in accumulated capacity by 2015. In other words, if in 2009, Europe had nearly 50% of all installations worldwide in 2014 these represent less than 35%. Before 2020, with the current trend, Europe will have less than 30% of all wind turbine installations, unless offshore wind installation really “takes off”, after a very gloomy start. In terms of wind electricity production, europe is even further behind China since many installed turbines are very old (i.e. inefficient and small in capacity). Consequently Europe generated “only” 257TWh, less than the amount of electricity from wind generated in China.
The EWEA initially estimated in the 2009 (see the “Pure Power scenario” paper) a total of 230GW of installed wind capacity by 2020, producing 580TWh of energy per year, with 40GW coming from installed offshore wind power. However, new scenarios drafted between EWEA and main sector players in 2014, now point towards 192GW of installed capacity and 442TWh of energy by 2020. Even for this reduced scenario to be achieved, between €90 and €100Bn will have to be invested until 2020 a very considerable investment under the current economic conditions. European turbine producers have approximately 90% of the European market, roughly 30% of US market and less than 10% of the Asian market. European manufacturers Vestas, Enercon, Gamesa and Siemens Wind represent nearly 35% of the whole market share of wind turbine producers but Chinese players are also over 30%.
As anticipated by this consortium, in the proposal submitted in 2009, the maximum size of wind turbines has not changed significantly since 2009. Enercon’s E-127 (currently in 7MW form) remains the largest commercially available turbine. Turbines of 10MW are in development, but these are being tailored for off-shore and will likely only be competitive in “extremely favourable” off-shore wind park environments/scenarios. The evolution of the average turbine size in Europe, in the past 5 years, shows signs of an increase (i.e. total installed capacity divided by the number of installed wind turbines), but this trend does not seem sufficient to fundamentally change the most significant argument, presented by this consortium, in the 2009 proposal: wind turbines face challenges to grow past the 6-8MW size and will be significantly more expensive in the 20MW range, especially if no good offshore locations, to install them, exist!
If conventional wind power has been mostly stagnating in Europe since 2009, as has been described above, the same cannot be said about Airborne Wind Energy (AWE) and HAWE (High Altitude Wind Energy) technologies. It is true that market status has not been achieved so far: currently, no AWE/HAWE is in advance-demonstration phase (i.e. in continuous operation for over 8000 hours, roughly 1 full year) but some industry trends are clear:
• A distinction has emerged between AWE and HAWE;
o AWE seems to have centred around operations without a turbine tower and an operating altitude between 150-300m based on concepts of cross-wind flying kites;
o for HAWE, concepts are less “standardized” (Magnus-effect based concepts coupled together with wind mill ladder type concepts, etc…) and operating altitudes around 400-800m have been experimented with.
• Major topics, in the industry (i.e. both for AWE and HAWE) now focus on:
o R&D metereological assessments, led by Ken Caldeira and Cristina Archer;
o Air traffic implications of AWE/HAWE especially for installation in developed countries where most of the airspace is either blocked or under severe usage restrictions;
o Control of the devices, especially cross-wind kites which will fly at above 50 knots in directions perpendicular to the main wind direction
• Providing demonstrations of capabilities, such as
o Makani Power in 2010
o Altaeros in 2013
o TUDelft (Laddermill Kitepower) in 2012 (700m of altitude operation)
Although it is hard to estimate, over 2000 professionals are currently working in AWE/HAWE concepts of or related work, with over 40 institutions currently in charge of this development. At Omnidea, and as part of this project, approximately 8 jobs were partly (or almost fully) funded during the duration of this project. Overall in the consortium, between 20-25 were almost permanently allocated to this project as per Figure 4 1.
At the current state-of-play, it is wise to introduce here the words of PJ Shepard co-founder of the Airborne Wind Energy Consortium:
“the progress toward working, power-generating, AWE prototypes has been impressive in the last few years. Some companies have proven energy-generation capability, many have demonstrated some level of autonomous control, and a few are planning power farm development”.
But, in the words of the industry: “scaling up past the prototype stage won’t be easy”. Experts in the industry believe that without strong government support, installing even one gigawatt of airborne power could take 20 years or more. Some larger companies such as Honeywell and 3M have shown some interest in the AWE/HAWE technology, but the major traditional wind power developers — Siemens, GE, Vestas, and others — have so far left the airborne designs to the startups.
A recent study by Cristina Archer shows that there are enough areas of the globe where consistent, 10m/s wind conditions exist for at least 15% of the year’s time. From this it is clear that the potential exists for AWE to provide generation power enough for Europe’s needs.
Another study from Colm O’Gairbhith from Loughborough University, published in 2009, compares wind availability at several possible test sites in Ireland and presents a comparison between typical wind speeds at 150-200m and wind speeds at 750m. A rough comparison gives a potential power increase of nearly 101% i.e. roughly double that at lower altitudes for typical wind speeds of around 9,16m/s (at 150m) and 11,86m/s at 750m. These values are based on the equation where wind power, P, increases with the cube of wind velocity, V. The same study points towards a capacity factor of 52% for kites which effectively work in a similar pattern to that of HAWE (a pumping cycle with a reel-in phase). Therefore, assuming that HAWE could also have a 4:1 ratio between the traction phase and the reel-in phase, HAWE could also have a capacity factor of 52% but at the it moment seems to be a little optimistic for HAWE but within the overall capability of the system). This means that there is approximately double the potential energy to extract (101% more) and a roughly 12-20% greater capacity factor (vs. respectively 40% capacity factor for off-shore turbines and 32% for on-shore turbines). Combining both factors it seems realistic to say that power outputs can triple from the use of AWE/HAWE technologies.
According to a report from Carbon Tracking, AWE/HAWE technologies have attracted the attention of the science and engineering communities, with articles having already appeared in the New Scientist, Scientific American (who identified AWE as one of the top five “Greentect” technologies of the future) and the New York Times. The Norwegian state-utility Statkraft, the largest renewable energy company in Europe, has identified Airborne Wind Energy as an energy technology for the future and has invested in technology development. A study, published in 2010 and entitled “Assessing the Potential for an Airborne Wind Energy test-site in Ireland”, presents a short overall description of the “by then” status of technological evolution as well as putting forth 3 potential test sites in Galway and Offaly counties. The test site assessment also presents results from a survey conducted with several identified AWE/HAWE technology developers, such as Makani, Joby energy, Altaeros, etc.. The study provides two budget estimates for site construction: one called “full costs” and another called “start-up” version. Costs vary between 2,5Mn€ (full cost) and 0,9Mn€ (start-up costs). This is very helpful in shedding some light into the benefits of using the Ota test site for free.
In March 2011, Omnidea was contacted by the then Makani power CEO, Corwin Hardham on the possibility of a partnership for Airborne Wind Energy. This possibility was discussed at length by email and a meeting was held at the 2011 Leuven AWEC. Nevertheless, because Makani’s prime interest was in a concept employing wings (which HAWE exchanged in favour of a rotating cylinder concept for the Magnus effect) the potential for future cooperation was judged to be small. On October 23rd 2012, Mr. Corwin Hardham died unexpectedly from cardiac insufficiency; Omnidea’s team hereby honors his memory as one of the AWE community pioneering minds; with Mr. Hardham having himself been the initiator of the talks between Makani and Omnidea, the cooperation between both institutions did not develop any further. In August 2013 Makani was acquired by Google.
During the 4th HAWE Progress Meeting, hosted by EDPI in Lisbon in March 2012, discussions were held with Prof. Uwe Paulsen, from DTU-WE regarding the possibility of exchanging information between the FP7 project DeepWind, coordinated by DTU-WE and HAWE. This possibility is especially interesting regarding the availability of aerodynamic coefficients for water immersed cylinders under a rotation motion, and its importance to the estimation of drag coefficients. Prof. Paulsen has since sent to Omnidea a paper describing the operation of the Deepwind project, as well as its challenges and accomplishments in greater detail. This collaboration is still on-going, especially regarding the estimation and calculation of 3D lift and drag coefficientsfor the Magnus effect.
Even during the proposal elaboration stage it was already clear that HAWE is supported by IPR, in this case this technological development is supported by the PCT patent “Atmospheric Resources Explorer” (PCT/PT2007/000022) which forms the basis for this consortium to undertake work in this Collaborative Project in FP7”. This was already referred to in the proposal and the PCT patent was finally awarded in December 2010, just 3 months after the beginning of the project. The “Atmospheric Resources Explorer” patent is the crucial IPR piece in this technology as it describes both the layout of a typical HAWE system (such as the one tested at Ota, i.e. with an hybrid-LTA ABM, a tether cable and ground station containing a winch) as well as the operation method of HAWE (i.e. through a pumping cycle).
Another patent, obtained at national level and under evaluation at international level, named "Airborne Platform" was submitted, at international level, on the 6th month of this project (i.e. March 2011). This patent describes in detail the ABM concept and design, aiming at protecting industrial property over Omnidea’s currently in development part of the HAWE technology. The invention pertains to aeronautical enginering and consist of an airborne platform that can be built to large sizes without requiring rigid structures of comparable dimensions and which uses both buoyancy and aerodynamic magnus effect for lift.
A third patent, also obtained at national level and under evaluation at international level, named "Submerged Platform" was submitted at international level, on the 19th month of this project i.e. (in July 2012). This patent describes in detail a system under traction stress (such as the mooring platform of a HAWE concept, which means that the ground station can be positioned below the surface, thus not exposed to adverse weather conditions with an exit emerging above the surface to allow tethering of the ABM.
As consequence of the work done by Lankhorst on this project, a patent application is being submitted for the new designed cable. Having as provisional name “Multifunctional Tether”, it will likely still be submitted in the beginning of 2015, likely both as an European/international patent and also in the US. Lankhorst’ push comes from multiple commercial inquiries for an electro-mechanical cable, which, despite will very likely materialize, into a real order, in the near future.No expenses have been presented, by Lankhorst, to the EC, on this matter (i.e. IPR registration). No expenses, relating to this patent, have been (or will be) presented to the European Commission during the HAWE project.
The HAWE website has been online since the initial Kick-Off meeting in Lisbon (Oct. 2010) and a new section for partners was added at the time of the Dubrovnik meeting. The site can be accessed through http://www.omnidea.net/hawe/ and partners (as well as the commission) can access specific contents through the “login” button. The HAWE website was the subject of an update in June 2012 and again in September 2014.
A moodle platform for file sharing was also created in June 2011 but it showed little to no acceptance by the project partners. As a consequence an ftp website, for data and file transfer, was created in September 2011 and has since been used by the partners. Through this ftp website, videos are also available, from the testing activities at Ota. Finally, since 2014, Conferences, articles and publications
Mr. Pedro Silva and Mr. Tiago Pardal attended the Airborne Wind Energy Conference 2011 held in Leuven, Belgium in May 2011. At this conference Mr. Tiago Pardal made a presentation “Omnidea’s system for HAWE-FP7” on behalf of co-authors Mr. Rei Fernandes and Mr. Pedro Silva which is also accessible through the HAWE FTP website.
At the 3rd HAWE meeting, held in September 2011 in Dubrovnik, Croatia, Neven Duic (FSB), Mr. Luka Perkovic (FSB) and Mr. Pedro Silva (Omnidea) presented a joint FSB-Omnidea paper on the CFD simulations performed during the HAWE project, entitled “Harvesting High Altitude Wind Energy for Energy production: a feasibility study of Magnus effect concept”, at the SDEWES Conference, on Sustainable Development on Energy, Water and Environment Sytems.
Also in September 2011 at the same conference, Prof. Josko Deur submitted an article related to the energy storage solutions possible to implement for an application such as HAWE in which a relatively high number of storage cycles for relatively low energy requirements but with relatively high discharging power requirements, are compared for the virtues of purely mechanical storage (i.e. a flywheel based solution) versus an electric solution employing supercapacitors.
In June 2012, Prof. Tatjana Haramina (from FSB University Department) presented a paper at the “Energy and Materials Research Conference” held in Malaga, Spain entitled “Permeation of hydrogen and other types of buoyant gases through foils for balloons and their tensile properties” from her work on diffusion properties of the inner film layers performed for HAWE together with Mr. Pedro Lima (Omnidea), mentioned as co-author.
An article has been published by N. Duic, R. Penedo and L. Perkovic for the “Energy journal” (Elsevier) titled "Estimating the spatial distribution of high altitude wind energy potential in Southeast Europe". It focuses on the development of a tool to find available spots to deploy an High Altitude Wind Generator based on the population density, the existence of roads and railroads, and other significant obstructions to the use of high altitude devices. For this study HAWE was used as the base reference for the assessment in the Southeast Europe region.
HAWE’s most important major dissemination activity was the participation in the Springer Verlag book “Airborne Wind Energy” published in 2013 by the editor Professors, Roland Schmehl (TU Delft), Uwe Ahrens (Hochschile Heilbronn, NTS Energy systems) and Moritz Diehl (University Leuven) to be part of the first comprehensive publication on airborne and high altitude wind energy. The HAWE concept makes up the 29th chapter of the mentioned book as a case study. For this publication, detailed HAWE simulations were produced, along with a detailed concept comparison with typical kite-based solutions. As a consequence of the above publication HAWE was also present at the 2013 AWEC (Airborne Wind energy conference), held in September 2013 in Berlin, through both a poster presentation and an oral presentation by Mr. Tiago Pardal, Mr. Rei Fernandes and Mr. Pedro Silva.
A final dissemination activity was undertaken in October 2014 when the Portuguese TV channel SIC (http://sic.sapo.pt/) undertook an in-depth news article about HAWE, broadcast on 14.10.2014 at the main SIC channel (on the evening news) and which can be retrieved at the following link (http://sicnoticias.sapo.pt/programas/futurohoje/2014-10-15-Usar-o-vento-para-produzir-energia-mais-barata-). This news item is part of the program “Futuro Hoje” (the future today) of which Omnidea’s technology represents one chapter (this program features, on multiple occasions, green energy technologies). This article was produced during a testing day at Ota Air Force base.
List of Websites:
http://www.omnidea.net/hawe/
Contact details:
Nuno Fernandes
Omnidea, lda
Room 2.1 - Ed. VIII
Campus da FCT
2829-516 Caparica, Portugal
This report describes a technology for harnessing energy from high altitude wind through a pumping cycle, in a vertical trajectory, executed by a hybrid lighter-than-air tethered rotating cylinder, which generates dynamic lift through the Magnus effect.
The pursuit of high altitude wind energy has led to a number of different systems and configurations which try to harness energy from the greater power densities available at altitudes higher than those used by conventional wind turbines. The HAWE consortium, comprised of 2 research institutes, 3 large private companies, 1 air traffic management authority and a Portuguese innovation-led SME coordinator, has developed, over the last four years, a system based on a hybrid airborne platform, which through the Magnus effect, converts wind power to force and speed (mechanical power) on a tether cable; as the platform executes a pumping cycle trajectory, mechanical power is converted into electricity, at a ground station. The concept is based on a PCT patent granted to Omnidea in 2010.
After initial trials with an airfoil in 2007, a rotating cylinder was selected as the lift force source for the Airborne Module, allowing the team to take advantage of the greater values of aerodynamic lift and drag available from the Magnus effect. Since 2011 the concept has been refined and further developed by the consortium, focusing on cycle modelling and optimization, as well as full system design and study. A 16m long by 2.6m wide, cylinder shaped, scale prototype, estimated to generate 3kN of force under wind speeds of 7m/s, was built at Omnidea and tested extensively at Ota Air Force base, where proof-of-concept was obtained, under operating restrictions, on Q3 2014.
This wind energy producing device is aimed at lowering the current wind generators benchmark costs (typically twice as costly to produce electricity vs. conventional fossil fuel sources) by aiming to make it competitive with coal derived electricity costs. Although the small-scale prototype could not confirm the significant cost reductions envisaged, the economic feasibility study conducted confirmed that, under the right operation parameters, the technology can be competitive with technologically mature wind turbines. Further to this, the aeronautical safety operation assessment evidences that, despite current air traffic regulations not being adapted to permanently moored/cycle operating platforms, delimited and identified demonstration parks can be installed if outside the influence area of airports and large cities. A detailed study for off-shore operation demonstrated the suitability of the concept to be installed, in deep waters and on “shore distant” locations, likely through the use of adapted tension leg platforms. This development also saw the issue of a patent application within the period of the project.
During the development stage, challenges where identified and solved. Still, the results achieved, at sub-system testing, uncovered components where even the current technological state-of-the-art must evolve: the multi-layer, multi-function Dyneema tether cable must improve to achieve a greater bending fatigue resistance, while different production techniques must be experimented with, to allow insert energy transmission copper conductors in a structural cable subject to millions of cycles per year.
Project Context and Objectives:
Currently all produced wind energy comes from Horizontal Axis Wind Generators, known colloquially as wind turbines. Modern wind turbines use lift force, generated as the incoming air is deflected around the turbine blades, to rotate the blades around the housing hub. Turbines incorporate features such as pitch mechanisms and yaw drives to increase extractable wind power, but are subject to the Betz limit (from the afore mentioned wind deflection) which limits maximum extractable power to 59% (16/27) of theoretically available wind power. But, as the turbine operating efficiency varies between 60 and 75%, in reality only around 40% of the wind power density available is translated into electrical power.
Commercial turbines on the market range from 0,5 to 6MW and start rotating at wind speeds of 2,5m/s, only being shut down if wind exceeds 25m/s, even though their rated (nominal) power is attained with wind speeds around 12-13m/s. Nevertheless, since constant winds at hub height are usually much lower, at around 6,5 to 8m/s, power production is only around 200 (minimum on shore) to 400kW (maximum off-shore) per installed MW. In June 2014 over 336GW of wind power were installed worldwide, and estimates for power generation, for 2014, should be close to 1 PWh (1 Peta Watthour) of electricity, representing 4-5% of all electricity used worldwide .
In trying to increase wind power production, the current wind tower technology has been aiming for ever higher hub altitudes, due to the existence of a greater power density in wind (i.e. wind blows stronger and more constant) but also because this allows for longer blades (i.e. a greater swept area). This poses an increased challenge in the towers, as the mass of these increases substantially with height. The engineering challenges that this quest presents demand an enormous financial investment in R&D and construction.
Airbone Wind Energy harnessing devices, differ from conventional wind turbines in that they extract energy from wind without the need for a turbine tower. Instead this is usually replaced by a tether cable connecting an airborne installed device to a ground station. The airborne installed device can extract wind energy:
• directly, doing the conversion to electricity in the airborne device, sending electricity through the tether cable to the ground station, or
• it can produce mechanical power which is then converted to electricity at the ground station, from the movement-force “pair” reaching the ground station through the tether
In what respects to the airborne device, it might produce energy:
• constantly (i.e. “always on” or always producing) or
• intermittently (for example through the use of a pumping cycle, similar to reciprocating engines, where a phase of energy production is alternated with energy spending to bring the cycle to its original, i.e. “start” position)
In what regards to the construction nature and principle of operation, the ABM can be:
• an aerostat, relying solely on lighter-than-air technology to stay aloft;
• an aircraft, being heavier than air and employing aerodynamic lift to stay aloft;
• or a hybrid, employing both technologies simultaneously.
Airborne Wind Technology represents a derivation from the conventional wind power development, with a multiplicity of technological options. Airborne wind harnessing systems have multiple advantages over conventional wind turbines:
• no need for a heavy and expensive turbine tower, which also means foundations are much less problematic;
• possibility to install the electricity generating equipment on the ground;
• much higher wind energy, from the higher velocities at altitudes above 1-2km, means there is no need to focus exclusively in optimization and efficiency;
• possibility to operate at different altitudes to increase capacity factor to 60-70%.
Additionally, airborne wind energy, aiming to attain high altitudes (above 1000m) is no longer influenced by the orography of the terrain over which it is installed, but rather subject to the availability of winds in the lower part of the troposphere. Potential seems to exist for significative cost reduction in electricity from wind energy, which will bring it in line with electricity generation from non-renewable sources (inc. coal). Nevertheless, as of 2013, apart from some demonstration parks, no such system was in commercial operation.
The HAWE project, was a 42 month, R&D centric technology development project, aimed at providing proof-of-concept and TRL increase of a hybrid airborne platform operating according to a pumping cycle arrangement to extract wind power and convert into electricity at a ground station, aiming to harness clean and renewable energy from high altitude wind.
The technology investigated consists of a lift generating airborne module (ABM) comprised of a rotating cylinder that is connected to a ground station by a tether cable, operating in a two phase cycle. During the power production phase the airborne module generates lift through the Magnus effect of a rotating cylinder, pulling up on the tether cable and converting the lift force into electricity at the ground station through an electrical generator. When the ABM reaches its limit of operation, the recovery phase begins as the rotation of the cylinder is stopped and the cable is reeled back to its initial position completing a cycle. The design and development of the three main components of the system, the ABM, cable and ground station involves a complex process with multiple interdependencies calling for several iterations to arrive at the design specifications needed for the development of the various components. The initial design capacity selected for the proof-of-concept demonstrator was a unit of 50kW, but for pratical reason this ended up being scaled down to a 10kW unit, in order to minimize time wasting efforts and focus on the development sub-system critical components.
The intermediate objectives were, therefore, to undertake an iterative study, design and implementation of sub-system critical components which could be individually tested before final assembly into the system (i.e. demonstration prototype). At sub-system level, the specific objectives were to design, manufacture and test:
(i) an ABM with the required aerodynamic characteristics, using inexpensive materials to provide a strong, gas-tight envelope that is UV resistant and capable of operating in typical wind environments. The system should have a low cut-in speed (1-2m/s) and a high cut out speed (20-25m/s) operating at altitudes greater than 200m;
(ii) a ground station consisting of a combined traction winch and electrical generator combining very high cable velocities with significant heavy loads and extremely long operating life spans, to perform a pumping cycle of the ABM for electrical power production, as well as the ground mechatronics layout and control software to optimize the operating strategy;
(iii) a non-metallic structural tether cable to secure the ABM, ensure data and electricity transfer between ABM and ground station as well as supply buoyancy gas to the ABM, whilst being able to withstand the stresses and pulling forces experienced in the operating cycles;
(iv) interfaces between the cable and the ABM consisting of swivels (one in the cable to avoid excessive cable twisting and another on the cable to winch interface, to always ensure the ability to transmit gas, electricity and data to the ABM.
HAWE was organized in 8 work packages plus a WP for management (WP90). WP 10 focused on the ABM design and construction, aerodynamics and main cycle simulation. WP 20, was essentially control related, featuring the layout of hardware for ABM control definition and selection of individual components as well as cycle optimization and control.
Cable design, including layout of the multifunctionalities (gas transfer, electrical power upload to ABM and data transfer and through cable communication to ABM), construction, testing and iteration, plus possible “large scale” manufacturing optimizations was the focus of WP30. Ground station development, from a design/construction point of view, was the main aim of WP50, which focused mainly on the winch/generator hardware part, plus the ground station control module. It run in parallel with WP60 which focused on energy storage and drive arrangement optimization at ground station, as well as components selection and control strategies for the control cabinet. WP40 centred on subjects related to more than a single module (whether airborne, cables or ground) featuring adaptations required to make the project economically viable for off-shore installation, aeronautical safety, environmental criteria for demonstration site selection and the definition of operational limits of the currently devised system. The WP also included definition of the cable swivels needed to avoid “in air” significant cable twist. WP 70 was dedicated to all testing (weather individual component testing) and the final proof-of-concept (i.e. on site, full assembled prototype testing) which ended up being performed at Ota Air Force Base, in Portugal, roughly 45km NE of Lisbon. WP 80 was dedicated to economic feasibility analysis, comparing an improved and scaled-up HAWE system (evolved from the prototype tested in WP70) with current state-of-the-art wind energy turbines to assess its economic competitiveness.
The HAWE consortium was arranged around the division of tasks presented in Figure 2-3:
Project Results:
1. Rationale
The main reason for embarking in a project like HAWE is that the energy return, from high altitude wind, must be sufficiently “rewarding” that the development of a completely new system makes sense. In latitudes similar to those of Europe (i.e. 36 to 58º of latitude North) the available wind energy at, for instance 2km, is by at least a factor of five times higher than that of typical wind turbine operating heights.
Wind potential grows with altitude, because wind speed typically increases with altitude, at 10km, for example, there is approximately 15 times more energy density than near the ground, with a very steep wind potential growth between altitudes of 3 and 8km. It is also of importance to understand what the characteristics of this wind availability are. As has been seen above the available energy contained in wind depends essentially upon:
• Latitude in the globe - wind is stronger between the 25º and 55º parallels, independently of the hemisphere chosen;
• Season of the year - wind is always, on average, stronger in the winter and weaker in the summer;
• Hemisphere - for comparable latitudes and seasons of the year the southern hemisphere displays stronger winds than the northern one.
Ken Caldeira and Cristina Archer’s research work shows that in the troposphere it is always advantageous to go higher in altitude even if also highly dependent of the latitude. For example, although at the same latitude, the northeastern coast of the US and the north of Italy, present different wind potentials for the wind speeds available in 95% of days (i.e the 95th percentile winds).
2. Concept description
The pursuit of high altitude wind energy has led to a number of different systems and configurations to harness energy from the greater power densities at altitudes higher than those used by conventional wind turbines. Since 2006, and with increasingly higher investment and technology readiness level, a system based on a hybrid (deriving lift from both aerostatic and aerodynamic forces) craft, tethered to the ground by a cable, has been in development at Omnidea. The lighter-than-air (LTA) structure takes advantage of the high aerodynamic coefficient of lift made possible through the Magnus effect, hence making this module a hybrid platform which we call the airborne module (ABM). The Magnus effect aerodynamic lift is generated by spinning a cylinder, allowing the system to take advantage of much greater values of aerodynamic lift and drag coefficients, when compared to conventional airfoil theory.
The system accesses wind resources at higher altitudes, by transferring the resultant forces of this wind power, through a multi-functional tether cable (structural cable including within it the functions of data transmission, electricity delivery to the ABM and buoyancy gas replenishment), to a ground station on the surface. The traction power of the ABM is used to produce electrical power through a pumping cycle, executed essentially in a 2D vertical plane in four phases:
1 – A power production phase (rising phase) in which the ABM is put into its maximum spinning velocity until the predefined operating altitude is reached; during this phase the cable is pulled and unwound from a winch drum coupled to a generator and therefore produces electrical power.
2 - A transition phase in which the rotation of the ABM is stopped as the maximum operating altitude is reached, resulting in a reduction of aerodynamic lift to zero, thus reducing the amount of energy required for recovery.
3 - An energy spending phase (recovery or winding phase), in which an electrical motor at the ground station rewinds the cable to its original position, bringing the ABM down to the starting position of the cycle. During this phase only the aerodynamic drag and the aerostatic lift of the ABM have to be overcome, but energy is still required to achieve it.
4 - As the minimum altitude for the cycle is reached, another transition phase takes place, with the ABM being accelerated into a spinning movement to begin a new power production phase.
The HAWE concept should thus be described as an airborne wind energy generator, capable of extracting wind energy from high altitude winds, through usage of a tethered, hybrid, airborne platform operating in a pumping cycle arrangement.
3. Component development
HAWE is mainly made of 3 separate (but interconnected) sub-systems:
• Airborne Module (“ABM”), comprising the whole flying platform, including the LTA balloon and envelope surface, the electrical hardware required to spin the LTA platform including motors and drivers, plus the hardware required to control the platform and communicate with the other systems (usually known as control box)
• The tether cable, including the functionalities of gas transfer from ground to ABM, electricity transfer media (i.e. electrical conductors), optical fibre for data transfer and the structural strength member itself
• The Ground station, which includes a winch, a sheave, a control hardware box with Graphical User to Machine Interface, plus the energy production and storage components (e.g. electrical generator plus a set of batteries, super capacitors or a flywheel)
The HAWE design is based around the following fundamental principles of operation:
• all individual components must be transportable in standard cargo containers,
• the concept must allow for modular (stacking) power increases.
• expectable lifetime should aim for 20 years
• the operating altitude must be changeable, at any time,
• low cut-in speed (2m/s) and high cut out speed (20m/s);
• The system will not operate under thunder storms, in which case it is reeled down and idled.
4. The Airborne Platform
The High Altitude Wind Energy (HAWE) production device is a machine designed to extract power from the wind harnessing available power at much higher altitudes than conventional wind turbines reach, or are likely to reach in the next 5-10 years. The task of designing a lighter-than-air platform that can operate at altitudes ranging from 200m to 2Km, with these altitudes changing every few minutes and with continuous 24h/day operation is an enormous challenge in itself. Thus several iterations were reqired to arrive at a stable and feasible prototype for testing in altitudes of up to 600m.
The main characteristics of the HAWE system are:
• High aerodynamic coefficients, made possible by the Magnus effect, which allow a high loading of the ABM even at low winds speeds of 2-6m/s without crosswind operation. To achieve these coefficients, the cylinder is spun faster than the incoming wind using an electric motor in the ABM.
• Lighter-than-air (LTA) design, enabling the ABM to remain airborne under conditions of no wind and stable under most wind conditions.
The study part regarding the ABM centered, mainly, on the following 4 areas:
• Dimensions trade-off for the ABM envelope geometry
• Aerodynamic study for the coefficients of lift, drag and moment calculation
• With all of the above, initial cycle calculations were performed
• Materials were also studied for the ABM
The altitude operating range mandates a certain pressure and temperature gradient between the interior of the ABM envelope and the atmosphere; it is thus the definition of the operating altitude range which determines the required strength of the ABM envelope, all layers included. The calculations of the internal buoyancy requirements are also not easy since they are dependent on multiple factors. Most of these calculations have variables which depend upon the definition of the operative requirements (cycle max. height and max. length, max. sustainable wind speed, etc…) but others must be estimated before the cycle begins (e.g. safety factor, etc…).
Due to the complexity of the calculations the team decided by a mix approach; a sum up of the expectable sub-system weight was performed, together with the expectable envelope and cable weights. After this was determined, specific dimensions were settled upon so that construction could begin; whenever new data would change the original project specifications, then the operational requirements would suffer from these changes. Furthermore, the system always requires a minimum relative pressure above 1.5KPa (i.e. 1500 Pa of overpressure vs. atmospheric pressure) in order to maintain shape and spin without major problems, therefore avoiding that its shape, when rotating, creates a “banana-like effect” around its center section, which induces multiple problems during the rotation.
Regarding volume-to-surface considerations, the ABM mass, for a given volume, varies with the cylinder’s length according to a relationship such as (1/l), while the weight of the ABM goes upwards as the l/d ratio is increased, as expectable by the variation according to (l/d). In summary a balloon with a high volume-to-surface relationship is desirable both to increase initial buoyancy (i.e. lower ABM mass for a given volume) as well as to maintain available buoyancy (i.e. reduce buoyancy gas losses). Still, in order to decide the diameter-to-length ratio of the test prototype, further studies were needed also, because it is the intent of this project to understand if the solutions being employed can be scaled for much larger ABM’s; hence several “typical” L/D ratios were investigated, such as 6, 8 and 10 (4 was deemed too small); for an L/D ratio of 6 full calculations were performed.
An extremely complex subject, regarding the ABM, is how to solve the huge forces, acting upon the ABM to gain volume as it goes up in altitude, towards a region of lower atmospheric pressure; this pressure gradient creates huge forces on the envelope, and must be countered by a Dyneema net, installed as the final 3rd layer of the ABM envelope. Nevertehless, due to the huge dimensions of the ABM, even Dyneema will reach its limits quickly, and hence a 1500-2000m maximum ceiling should be established. The problem increases proportionally with altitude, but becomes exponential with increased diameter; this is a further reason to keep a relatively low diameter, which in turns increases length for a given internal volume (i.e. buoyancy) requirement. Finally, there is the additional factor that small L/D will imply a high induced Drag which is something to be avoided as it reduces aerodynamic performance.
The materials selection is of paramount importance to the HAWE project as the ABM amounts for significant part of the cost and is the most likely to require frequent maintenance (due to exposure to an external environment). From earlier studies at Omnidea, the envisaged structure for the ABM should comprise 3 layers:
• a gas barrier layer or inner layer, where diffusion properties are most important,
• an outer layer, encompassing the inner layer, to withstand the pressure difference with the outside atmosphere without tearing, and to provide UV resistance,
• a structural net (a rope mesh) providing the bulk of the resistance and which is able to distribute the loads more effectively as, being a mesh, it can be oriented in a preferential direction
Candidate materials for the inner layer are thin film polymer combinations, basically arranged around a gas barrier plus bonding-friendly polymer arrangement (e.g. EVOH, a typical barrier film, being in a “sandwich” of 2 Polyethylene layers). Conclusions from the permeation tests induced the team to select a hybrid diffusion barrier material (PE/EVAL/PE) for the inner layer due to the low diffusion rates promised by EVAL, coupled with acceptable mechanical properties and good the bonding possibilities offered by the PE layers.
The outer layer materials are all of the polymer cloth type and, to join them together, one needs to stitch them. How easy it is to perform this task and how well both cloths are connected through this stitching, is what is mentioned as Ease of Work for the outer cloths.
A Dyneema structural net, with extremely high strength, was considered the perfect candidate for the outer net.
5. The Cable - requirements
The most fundamental input for the conceptual design of HAWE’s main cable is the correct perception of the amount of cycles required by HAWE’s operation. Each operating cycle is dependent from multiple variables, chief amongst which is the altitude interval of the cycle. In order to further proceed with calculations, a cycle altitude interval must be assumed as well as upward and downward speeds. But, with current knowledge it is fair to admit that, even for cycles, with altitude intervals as wide apart as 500 to 3000m (2500m of altitude difference) the upwards speeds should be assumed between 2-3m/s and downward speeds around 1,5x that, (i.e. 3-4,5m/s). Assuming an average of 2,5m/s of speed upwards and at a minimum of 60º of inclination, the “pure vertical speed” is around 2m/s. This means that 1250s are needed to go up in the cycle and 1250/1,5 s will be needed to come down, for a combined sum around 2000s. Hence, if a day has 86400s the minimum cycles which HAWE will perform in a single day is approximately 43 cycles. Likely HAWE will perform more, with an absolute minimum around 100 cycles per day more likely. But even taking 50 cycles a day this means over 17.000 cycles per year at 95% availability. It must be remembered here that HAWE’s specifications call for 20 years of operational life time.
To understand the problem further it must be estimated the amount of bending cycles, on the cable, per each HAWE cycle. This amount depends on the current layout of the apparatus and the driving winch. With the current layout we have 4 rope bending cycles, per each HAWE cycle, corresponding to bending on the storage drum and on the pivoting sheave, both on the upwards and downwards phase. This means 4 bending cycles per each power producing cycle, for a total of approximately 70.000 cable cycles per year. At 20 years of operational lifetime, this translates into a minimum of 1.400.000 cycles.
6. The cable - Structural part
There are several aspects to consider regarding the tensile properties of a fiber. Besides the importance of the tensile strength, it is also important to consider the fiber elastic elongation (related with the elongation at break). This last property is a measurement of the work (energy) necessary to deform the fiber.
Regarding cyclic bending-over-sheave (BOS) tests made for 3 material (steel wire, Dyneema and aramid) cables it can be seen that the Dyneema presents a linear behavior with a superior performance, when compared to the other materials under the analysis, so suitability of this material for loads of this typology is evident.
Initially Zylon appears as a good candidate but its strength is significantly impaired by exposure to high humidity/high temperature, and UV/visible light (precisely the operating conditions of HAWE) leaving Dyneema as a very good candidate in terms of abrasion, creep, UV performance and flex/fold resistance.
For the exterior abrasion protection jacket, the polyester fiber is the most suitable material.
7. The Cable- packaging
Traditionally, most cable designs place the conductor package in the cable center and apply reinforcement, around it. It is also common to add to the cable with a protective cover or jacket. In this layered construction the electric element is well protected from foreign object damage. A distinct approach is to place the strength member in the center (core) of the cable and arrange the conductor package around it. This construction has the disadvantage of placing the conductors on a less protected layer of the rope.
When the cable is put under load, the conducting core, the strength member and the jacket have to respond to the imposed tension. However each of these components has a very different stretch response. If the load imposed to the cable results into a higher extension than the said 0.45% elastic limit in the copper elements, the material will suffer permanent plastic deformation. After that plastic deformation, if the load is released, the material will only recover the elastic elongation, the excess conductor length has to be accommodated by the cable structure. If the cable structures acts like a compact assembly, the excess of copper length is forced to squeeze itself into a geometry that absorbs the plastic deformation length excess. The copper wire will form, at random locations, small double reversed bends. This behavior occurs each time the cable is again loaded, and will lead to ultimately failure, due to local fatigue effects.
For cables where the conductive copper elements are completely aligned with the strength member, only when the applied loads are very light, the copper can survive the loading cycles without plastic and fatigue issues. The conductivity of the copper wires is also influenced by the elongation of the wires. If the elastic limit is exceeded, the conductivity is altered and a more resistive area is created in the wire, which could lead to failure because of overheating and also damage the fiber strength member.
During the product development stage, several multifunctional cables concepts were developed, in order to integrate all of the required functionalities. For the development of this multifunctional cable, the low weight, diameter and high tensile strength were considered. Previous consortium member experience in the manufacture of high performance cables, indicates 12-strand braided ropes to the most suitable construction for the intended purpose.
For a HAWE system it is required a multi-function tether cable featuring a multi-layered cable with the following functionalities:
• A Structural “section”, for which breaking strengths, specific weight and fatigue performance are the crucial performance drivers;
• Power supply, to the Airborne Module, through copper electrical conductors running through the tether cable;
• Buoyancy gas exchange, between an high pressure He/H2 reservoir at the ground station and the ABM’s envelope
• data transfer, through optical fiber, for redundancy in communications
• The cable must also be possible to illuminate or be self illuminating
8. Ground Station - Winch discussions
From the multiple challenges that the HAWE concept poses in its development process, one of the greatest is perhaps the extreme operating conditions which the cycle execution, according to optimum aerodynamic parameters, dictates regarding the unwinding and rewinding speeds at the ground station, namely at the winch.
The winch system (that integrates also the control of all system setup) is the most important equipment at the ground station and, from an innovation point of view, one of the most relevant in HAWE. Per virtue of the magnitude of challenges considered above, the issue of developing an innovative winch was contemplated, at proposal stage, through an “innovation centered” approach in an initial delivered report where a trade-off, to reach winch optimum parameters (in terms of arrangement) is performed. The objective, through a component study interaction, is to find the winch arrangement which optimizes performance and is:
1. sufficiently rugged for everywhere, “always on”, operation for 20 years;
2. compact enough for transport without insurmountable logistic issues;
3. cost efficient enough not to compromise the overall project goal: undercutting current wind turbine energy production by up to 50%;
In a Winch the drum is the main component, and usually incorporates a power system interface (e.g. electric or hydraulic motor or generator/pump). The drum main dimensions, namely the diameter and its length are defined by the drum functions. The required drum diameter is calculated from the rope force and the power system torque, or from the rope linear and drum angular velocities.
Here, it should be noted that an additional restriction on the drum diameter is enforced by the rope durability. Namely, in order to minimize the rope damage caused by bending/straightening cycles, the drum diameter needs to be above a certain value, which is determined by the rope diameter and the rope material stiffness. Bellow the boundary value, the rope life expectancy diminishes significantly; hence from the rope lifespan point of view, the drum diameter should be as large as possible.
Once the drum diameter is determined, the drum length required to store the rope can be calculated. The number of rope layers is usually limited to 4 or 5 as a compromise solution, between the two opposite requirements (i.e lower length and longer life time), must be found.
In order to spool the rope onto the drum in a correct manner, i.e. without crossing the rope over itself or leaving empty spaces, the attack angle of the rope spooling (with respect to plane perpendicular to the drum main axis) must be kept at low values. Depending on the dimensions of the drum, the maximum allowable spooling angle requirement can be achieved without any additional special intervention. Due to the more complex motion of HAWE's ABM and a high probability of using multilayered drum (due to the long rope required), a guided spooling is necessary. The simplest solution to the spooling problem is to set up a separate pivoting sheave, placed at a given distance from the drum.
For HAWE’s constraints the sheave needs to be positioned at least 20 “drum lengths” away from the center of the drum. As an alternative, the pivoting sheave can be connected to the winch by means of a console, thus being rigidly connected to the winch. Single layered drums require no additional spooling guidance; however, multilayered drums usually require additional spooling mechanism to enable correct spooling of the rope in each layer.
An additional requirement for the winch intended for HAWE application is related to the special tether cable, which is to encompass power, signal wiring and a gas tube for ABM replenishment. Consequently, in order to connect the wiring and tube from the rope to the ground station systems, a special slip ring system needs to be used. Depending on the final system solution, at least one drum side has to be unoccupied by full shafts to place the slip ring system. Solutions using a hollow shaft with sufficient inner diameter may have certain advantage over the full shaft solutions, because the slip ring system can then be placed on the motor/generator side (far away from the winch), thus leaving one of the winch sides free for any other purpose (e.g. for double generator connection, or for using separate slip rings for power/signal and gas). Three different configurations were investigated by Rapp Hydema for the system drive choice:
1. The mechanical driven flywheel concept alternative; with good potential for future serial produced HAWE power generation plants. The cost impact and detailed adaptation of technology in such a system is unknown at this stage.
2. Several HAWE power generation plants working “out of phase” - this could present an alternative without energy storage requirements. To obtain a constant power generation to the net/grid, at all times, two (or more) systems must be in a power production phase per each system in the recovery phase.
For the proof-of-concept trials at Ota a direct pull winch, in a system without energy storage, was used employed, based on the following advantages:
• Will provide constant torque and consequently variable pull and speed versus the pulling diameter;
• If the drum is long, the difference in pull/speed on first versus top layer will be smaller;
• If the distance from the center of the drum to the fixed block is 20 times the drum length there would be no need for a level wind system (self-spooling);
• Will get 1 cycle (360⁰) at full load – not considering the level wind system;
• Limited mechanical contact between the rope and the drum for a winch without level wind;
• Physical contact between each wrap of the umbilical when spooling on to drum;
• Less complex design (in comparison with a traction winch);
The arrangement selected was a pivoting sheave as a separate unit plus a level winder. The HAWE system used a PTS Pentagon Research hardware control equipment.
9. Ground Station - control software
The ground station control is intimately related to the cyclic nature of the system as well as to the issue of energy storage (to "even out" the energy output to the grid, stemming from the pumping cycle nature of power production). To this end a report, including a comparative review and assessment of mid-size energy storage systems (< 10MW) was performed comparing flywheels, hydropneumatic accumulators, electrochemical batteries, and ultracapacitors, for criteria such as efficiency, costs, capacity and reliability. The study identifies ultracapacitor-based energy storage systems as the most viable for self-standing HAWE energy storage solutions, excelling in terms of running costs, acquisition costs, availability, efficiency, bulkiness and durability (i.e. system longevity) at low-to-mid altitude ranges of the ABM (i.e. when a relatively large number of charging/discharging cycles is needed). Further conclusions include the flywheel energy storage being competitive in the low-to-medium altitude range and sodium-sulfur and flow batteries becoming viable solutions whenever the number of cycles is significantly low.
With the modeling allowing calculation of the 2D cycle trajectory, multiple electromechanical configurations and energy storage systems, for the ground station, were simulated to achieve the optimum operating parameters regarding winding and rewinding speed as well as spooling speeds and torque.
The high-level controller manages the power flow to the energy storage subsystem, in order to provide a constant grid power flow over the whole HAWE cycle, while satisfying the condition on storage state-of-charge (SoC) sustainability. The high-level controller has been first developed for the case of electrical storage system, where the battery, ultra-capacitor and flywheel storage technologies have been considered. The controller includes feedback and feed forward actions, and a low-bandwidth filter used in the feedback path to average the SoC over the HAWE cycle.
10. Theoretical results, simulations and calculations
Since no calculated aerodynamic coefficients were available initially, to perform basic cycle performance calculations, Reid’s (1925) set of aerodynamic data for cylinders subject to Magnus effect, were used. The performed aerodynamic simulations provided coefficients for the cycle simulations and optimizations, thus giving insight on power production and recovery phase of a single module. Single cylinder was simulated, as well as two cylinders with varying distance between them.
The impact of turbulence closure on results is very large and requires estimation, which was achieved employing LES (Large Eddy Simulation) using the Coherent Structure Model (LES-CSM) approached. Results of aerodynamic coefficients for single smooth cylinders show a larger than expected amount of lift force, obtainable from cylinder rotation. Results for two cylinders indicated that extra lift can be obtained, but that small distances between the cylinders (less than 2-2,5 diameters) should be avoided due to a significant impact on the aerodynamic coefficients for each individual cylinder.
A “realistic” model, developed by Omnidea and DTU, accounting for factors such as:
a) the basic 2 Degrees of Freedom (2DoF) ABM model including 3 rope sub-models of different complexity,
b) formulation and implementation of HAWE system control variable optimization problem including preliminary results and insights,
c) the multi-body model concept intended for ABM passive stability analyses and more rigorous simulation studies including control system robustness analysis, and
d) an analysis of multi-balloon system-based control authority over yaw (and roll) dynamics.
was developed and further evolved, for a final iteration with the university of Zagreb. At its final stage it allows representation of a full cycle in 3D representation, as well as much more advanced cycle performance calculations. A real-time forecast system, capable of providing real wind conditions over multiple locations (proposed as potential HAWE testing sites), was implemented. Data was subsequently treated, through DTU’s forecasting software (WRF) to provide simultaneous measurements of time, height, free stream wind velocity, wind direction, temperature and air density and viscosity. This WRF modeling was implemented with Omnidea and has been updated to include several real wind speeds measurements at Ota test site.
To achieve a pre-commercial unit design of 80 kW of average power (roughly 200kW of peak power), based on the same simple concept of one ABM coupled to one ground generator, four rotating cylinders will be stacked in a single ABM. As the operational requirements mandate that the cable speed for the recovery phase is limited to 6m/s, this limits the maximum Cp that can be achieved to 1.235 (vs. 1.603 for a higher recovery speed). For simplicity of control and actuation, the cable speeds and the angular speed of the cylinder are constants.
The simulated cycle achieves peak power, at the end of the production phase, because that is the region with the highest wind speed. Therefore the average power of the generation phase will be lower than the peak power. For the cycle simulated, the peak traction power is 203kW and the average is 166kW. The total time for the generation phase is 568s. The average power spent to rotate the cylinder is 6kW, mostly due to inertia, whilst the power spent during the recovery phase, which lasts 398s, is 38kW. The average traction power for the whole cycle of 16 minutes is therefore about 80kW.
The energy produced in this cycle is over 5 times higher than the whole energy consumed (recovery plus rotation) and over 6 times (if recovery only is counted). It is expected that, with improved control strategies this value could be increased to 8. These simulations show that the components in the system need to be oversized to accommodate the fact that the peak power is higher than the average power by a factor between 2 and 3 (the factor for the actual cycle simulated is 2.52).
According to initial simulations performed at Omnidea for a single HAWE system, the ratio between the system’s maximum power production, achieved during the upwards phase of the power cycle, and the average power production per cycle (taking into account the transition periods, the energy spent to recover the ABM to the original position and the energy to rotate the ABM) can be up to 2,5; i.e. average power production is roughly 40% of the cycle’s peak power. This ratio influences the choice of components for the ground station and ABM, depending on the configuration of the power generating devices. Also, because the electrical grid requires energy to be fed in the most constant way possible, it is always desirable to “flatten” the energy feed-in curve; this curve is not flat at all for a single HAWE system, due to the pumping cycle’s nature. This then creates 2 issues:
i) If a single HAWE system is operated, then an energy storage solution is required to perform the ABM recovery, or a connection to the grid must exist;
ii) In the example previously presented, despite the production of “only” 80kW on average, some components must be dimensioned to accept over 203kW peak power and this is without considering factors of safety on the equipments.
Hence, compared to the basic configuration with a single HAWE system, alternatives (employing two ABM units) were analyzed:
a) wind farm configurations, by having multiple HAWE systems in a single location, with the main advantage being a reduction (or elimination) of the need for energy storage.
b) dual-winch with a single-generator configuration: this configuration allows installation of a generator unit more closely matched to the average produced power, significantly reducing the investment costs per unit of injected energy into the grid, compared to a single HAWE system configurations. The results can be better understood from Deliverable 50.3.
Apart from the basic, single-unit HAWE system configuration equipped with electric storage system, two more advanced configurations have been assessed for systems containing two ABM’s and two winches: dual generator configurations and single-generator configurations.
However, the single-generator system under a dual winch arrangement is only semi-controllable since there is little possibility of independently performing speed control for both winches; this leads to situations where the imposed winch speeds lead to non-optimal load-speed combinations.
Hence a third alternative has been proposed: a single-unit configuration based on a compact transmission arrangement with the use of direct flywheel (or hydro-pneumatic storage solution) for energy storage. This system provides a good blend of favorable overall performance characteristics with simple but enhanced controllability and reasonable investment costs. Similar structure of the high-level controller has been applied in the case of ground station configuration based on a compact transmission, including a direct flywheel storage connected to the winch through the gearbox. Here, the winch speed cannot be directly controlled during the ascending or descending phase (there is no low-level controller), because the winch speed is determined by the approximately constant flywheel speed. This can affect the ABM control performance, particularly for optimal performance control which happens when the ABM achieves near-vertical motion (meaning that there is almost zero drift with wind and thus no loss of energy due to decrease in relative wind speed). On the other hand, the transmission clutches are properly controlled (shifted) to provide smooth ABM reversing, with non-excessive clutch energy losses. In the case of planetary gear-based compact transmission, the winch speed can be effectively controlled by means of a low-level generator speed controller. This is because the planetary gear cinematically decouples the winch from the flywheel. The high-level controller has a more complex structure due to the more complex kinematic arrangement of the transmission.
11. Proof-of-concept and on-field results – test site preparations
Tests were conducted at Ota Air Force Base to be performed on days with the wind coming mainly from a “North oriented direction”, due to the specific needs to launch HAWE. The restrictions to operate under the predominant wind directions means that the entire length of the runway, (up to 2500m) is available as a safety margin in case of system failure, thus ensuring that whatever incidents occur with the prototype, it will always be confined to military facilities.
The assembly of the winch at Ota required some terrain preparation work such as ground leveling and construction of two concrete slabs. The main concern was to ensure proper alignment between the winch and the sheave. To satisfy the demands for electricity during operations at the field site a diesel generator was rented.
The HAWE concept operates at different altitudes during a single production cycle. It is therefore important to have reliable measurements of the wind conditions at the different heights of the Airborne Module. The wind data at Ota was provided by LIDAR providing wind profiles up to 200m and an ultrasonic 2D unit fixed at a 10m reference height. One of the objectives of having the two independent units to gather wind data is to be able to validate the forecast models developed by DTU and to assess the correlation between the 10m reference and the wind profile with altitude, so that reliable data on the wind profiles, without having the need for the expensive LIDAR unit, can still be obtained.
Prior to any regular tests or outdoor operations, a check list sequence was created to check proper functioning of all systems avoiding situations in which tests would be carried out without being able to, for example, record or log the obtained data. To assist these tests, validation of electromechanical components, gas-tightness of the ABM and confirmation of the communication capability between the ground station and the ABM was necessary. Also, the test site needed to be supplied with adequate mooring structures to assist the anchoring of the ABM. Additionally these structures could be used not only during pre-tests but also to support the phases of outdoor tests, in particular assistance in the takeoff and landing phases. Three support structures are required for operation:
• ABM trolleys to move it from inside the storage tent to the tests area in safety.
• Two small winches to allow for three mooring points for the ABM, before takeoff or after landing.
• Two metal framework supports to accommodate auxiliary cables from the small winches that constitute part of the mooring system of the ABM.
12. Proof-of-concept and on-field results – wind study at Ota
In order to have an idea on the importance of atmospheric parameters, besides the wind sonic station also a LiDAR equipment as well as the WRF (Mesoscale Weather Forecast Model) were used, both supplied by DTU-WE, to study wind at Ota. The LiDAR performs actual measurements, once every 18 seconds, for varying altitude levels until 200m while the WRF is only a forecasting tool, giving a forecast value once each hour and for a very broad swath area (nearly 1,5km2). The WRF can provide estimates until 2000m for 14 different height steps.
Low winds (< 5 m/s) occur during 55% of the time at an altitude of 10 meters and approximately the same at 15m; it is when we go up, towards 100m, that things change as this percentage drops to around 25% only. Between 100 and 200m the trend is for winds stronger than 10m/s to become much more probable; according to the WRF model, from 200 to 300m this trend is completed in the Ota test site as, all the way to 1000m, only winds in the 15 to 20m/s range clearly gain in probability. According to data provided by the WRF model for the Ota site, for altitudes between 500 and 1000 m, the WRF model predicts a significant increase in wind intensity, but also a shift in the importance of wind with the SW-W direction becoming more and more important as we near 2000m of altitude. This comes mostly from a decrease in the NW-N direction, while the W-NW starts strong, loses importance between 300 and 1000m and returns to its initial magnitude between 1300 and 1600m.
Analyzing the data, we see that, as we go up in altitude, the highest wind frequency peak moves to the right (i.e. towards stronger winds) as we go up in altitude. This can be considered the "normal" or "typical" curve format.
The corresponding cumulative annual wind energy from LiDAR measurements, shows that the energy available at 10m of altitude, 104m of altitude and 199m of altitude is respectively 0.5 2.5 and 4.7 MWh per m2 per year. Another way of interpreting these is to say that, at 10m of altitude, nearly 36% of all wind energy comes from winds below 5m/s, a percentage which decreases to 7% for 104m of altitude and even further (to 4%) as we approach 200m. Similar interpretation reach over 90% of wind energy from winds below 10m/s at 10m of altitude, over 80% at 100m and slightly over 50% at 200m. Above 200m there is no LiDAR data and, hence, we must rely on WRF data.
Thus, in order to get a better perspective on the agreement between both data sources, a comparison, at the highest altitude level available from both sources (200m) was made. The comparison reveals the issues with the WRF model: instead of having a total 8MWh/(m2.year)] predicted by the WRF model, the LiDAR measures no more than 4.7[MWh/(m2.year)] i.e.: only 60% of the value given by the WRF Model. We expect such a correction factor (60%) or even higher could have to be applied to the WRF data if a more conservative wind energy availability is to be computed for higher altitudes.
After having understood how the LiDAR and WRF data correlate, the histogram and cumulative histograms for the 300-1000m of altitude intervals reveals that as we go up in altitudes between 300 and 1000 meters the shape of the wind probability graphs does not vary significantly in this particular site (hence probably one of the main reasons why it was selected for take-off and landing of military aircraft), although until 500m the probability peak is still to the right of the probability peak at 300m. Nevertheless things are much less clear for 700 and 1000m where even "double peaks" can be seen, i.e. the probability function peaks twice, in both cases between 5 and 10 m/s.
This translates into the fact that, at least at Ota, although 600m is teh maximum reachable due to air-traffic regulations a system designed to operate, for example, up to 1Km above ground, would benefit more from the constant characteristics (smoothness) of the cycle, then from a clear increase in energy availability.
So how does the system compare with its theoretical maximum, in terms of energy? A simple calculation can be performed: applying the correction factor (between LiDAR and WRF) presented, and operating at 500m, HAWE could theoretically harvest around 8 MWh/(m2.year) (i.e. 13 MWh/(m2.year) x 0,6). This can be used to estimate that a prototype, such as the one tested with roughly 40m2, could harvest not more than 320MWh per year, or an equivalent energy production approximately 3kWh per each 5 minute cycle (i.e.: 36kW of power).
13. Proof-of-concept and on-field results - mechanical layout efficiency
The specifications for the ground station present in the field tests at Ota proving grounds were for an ABM with a maximum generated power during the production phase of around 60kW to achieve an average cycle power of 20kW. Nevertheless in testing with the current prototype the power figures were much smaller due to its much reduced dimensions, with a maximum of 10kW during the recovery (Ground Station electric machine working as a motor) and 3kW during the production phase (Ground Station electric machine working as a generator). Consequently and because there was clear evidence that the conversion of mechanical energy into electric energy, at the Ground Station, was not being done efficiently, a no-load test was performed.
For the no-load test of the Ground Station, the cable was completely reeled-in onto the winch drum and then the drum was rotated. Since the cable was not connected to any mechanical load, the only power consumption under constant speed was due to power losses of different components associated with the Ground Station system, namely on the winch gears, the driver and the resistor used to dissipate the energy produced, since the whole system is not connected to the grid. During the no-load tests the winch showed a power consumption of 16.4kW at 6m/s and 18.9kW at -6m/s (reverse direction). These power losses represent a permanent base load power consumption of 23% of the maximum power (60kW) just to spin the drum. Although this value is already quite high, an even greater problem occurs when operating far from the maximum power regime as, in terms of percentage, the power losses becomes much, much higher compromising any and all small net electric power production.
At this stage it was clear that there was a problem with one of the components subsystem, compromising the experimental potential proof of concept. Analyzing the data to analyse the efficiency of the Ground Station system, the Mechanical power and system efficiency can be calculated from the equation which states that the mechanical power equals the multiplication of the force (at the load cell in the end of the cable) by the cable speed.
The regime of operation of the Ground Station in the Recovery phase (electric machine as motor) spans approximately from 0 to 10kW with the efficiency achieving a peak of 50% around 9kW of mechanical input with a plateau of around 30% from 2kW to 8kW.
The regime of operation of the Ground Station in the Production phase (electric machine as generator) spans approximately from 0kW to 3kW. The problem is that, most of the time, the mechanical power input is not even enough to overcome the Ground Station power losses, so the installed winch would actually be working as a motor (i.e.: spending energy) when it should be working as a generator.
In the Recovery Phase (i.e.: system operating as a motor) for a given cable speed there is an increase in the efficiency with an increase of the mechanical power input as expected. Since power = force x linear speed, and there is a direct correlation between speed and voltage as well as between force and current, it can be concluded that the installed winch system will have higher efficiencies for higher cable forces almost regardless of the cable speed. The same analysis seems to be true for the Production phase. To support this conclusion it is important to understand that, for both phases, the efficiency is almost linear with the cable force and that the efficiency in the Recovery phase has a peak, at 80%, for a cable force around 4500N. Because, during the production phase the force required to overcome all the losses in the Ground Station is 1.5KN (i.e. only above this value of force is the installed Ground Station actually capable of producing electricity) then we can understand the reason behind the relatively small amount of electric energy produced: In the installed winch for testing, the efficiency is higher for low cable speeds and high cable forces, which is the exact opposite of what was needed in HAWE.
Unfortunately the threshold of 1500N (roughly 12.5% of the maximum load) to start producing electricity is too high to be overcome in almost any day of testing. Hence it became clear the components of HAWE’s ground station were not adequate for the typical operating regime of the prototype; more so, during the final season of testing the lack of ground station “suitability” and “efficiency” was the single largest contributor to not having been possible to execute cycles with net positive electrical energy generated. Nevertheless because the prototype was fully instrumented it was possible to achieve proof-of-concept by proving that there were cycles executed producing net positive mechanical energy. This net energy production is demonstrated by extracting the forces and speeds exerted on the tether cable (i.e.: mechanical power).
14. Proof-of-concept and on-field results - analysis results of a representative cycle
Some cycles were performed between the 23rd and the 24th of January 2014 and, cycle number 4 is the cycle which will be analyzed in detail:
-The duration of the cycle was of 70 seconds;
- the average cycle height was of 191m;
- the wind at average ABM altitude was of 6,23m/s;
In this case the mechanical power produced at the winch is around 1kW, despite a 0,35kW power consumption to rotate the ABM.
Cycle number 4 presented a rather high mechanical power, and the 2nd lowest wind at ABM height when compared to other cycles. Cycle altitude can be considered average as can the ABM power consumption when in comparison to other cycles. In this cycle, the cycle did not begin and end exactly at the same height; the cycle is considered performed when the cable length, at the end position, is the same as the initial cable length. Minimum cycle height was around 150m while maximum cycle height reached around 220m. The 3D trajectory plot shows a drift with wind of about 50m for an altitude difference of around 70m.
Cylinder rotation was kept constant at 50 rpm during the whole power production phase, being reduced to 10rpm during recovery. This means that recovery was not performed with a non-spinning ABM, but rather with a still spinning ABM, which increased the power needed for the recovery phase and, consequently, decreased overall cycle efficiency significantly. This is a necessity in HAWE in order to keep flight stability (i.e. otherwise recovery becomes very difficult and takes much longer with the possibility of crashes against the ground). Nevertheless, the main objective of the tests was to validate the simulation models and running the latter with similar parameters showed the same results therefore demonstrating that, under the proper parameters, HAWE produces net positive energy.
Cable recovery speed was around 2,2 m/s with an upwards cable velocity around 1,5m/s during unwinding (power production phase). Hence the “rule of thumb” of the velocity being 1,5x greater during recovery (vs. power production phase) was approximately achieved in this cycle, although both speeds were below that considered optimal. Power needs during recovery were around 1kW, however, at the beginning of the recovery phase there is an important power consumption with a peak of approximately 4kW. This peak significantly worsened the results and was due to the fact that, when recovery was started, the ABM was still rotating at nearly 25-35rpm due to significant inertia to slow the ABM rotation. Nevertheless, as previously mentioned, the results validated the theoretical model and the computational simulations.
The results obtained confirm the validity of Flettner’s work in spinning cylinder’s, dating from almost a century ago (1925). As expected the lift coefficient obtained was slightly below the theoretical values, with the same happening on the drag coefficient side. The measured CL/CD curve has an evolution which greatly resembles literature values with two minor exceptions. One of these being that the measured drag coefficient suddenly shows an erratic decrease and the other (towards the final phase of the cycle) the measured lift falls lower than expected. These two cases of measured lift being below “expectable” can most likely be traced to the misalignment of the ABM with incoming wind and by the system’s “elasticity”, measured by the cable catenary, which causes delays and fluctuations in the system force, both in intensity and in direction.
Throughout most of the upwards phase the operation is at a speed ratio, between 1 and 1,5. During recovery, as the rotation is reduced severely, the speed ratio falls below 0,3. It must be reminded that the HAWE system was designed to operate with speed ratios close to 2,5-3m/s during power production phase, with an aerodynamic optimum between 3 and 4. Nevertheless operation at these speed ratios requires the ABM motors to be capable of sustaining an rpm level around 90rpm or higher. This (90rpm ABM rotation) was achieved during the final phase of testing, after October but, unfortunately, the LiDAR equipment was no longer available at Ota.
The system model is a 2D model, which takes into account that the balloon is always aligned with the wind direction. We know that during the upwards phase, the tension in the cable ensures this alignment but, during the descent phase, tension on the cable decreases significantly, making the trajectory unstable and extremely susceptible to lateral displacement (sideways) easily becoming misaligned with the wind direction.
The magnitudes of the different powers which influence the HAWE prototype during the analyzed cycle are:
• Average power generation, during the power production phase is 2,4kW;
• Average power spent during the recovery phase is 1,4kW;
• The average power needed to rotate the ABM was of 0,35kW, leading to an actual “net” cycle power of 1,05kW (excluding transition period).
(Note: Transition period is always constant and has a high influence on average power depending on the length of the cycle)
These test values were obtained at 5,4 m/s of average wind speed (this average wind is the true average wind throughout the whole cycle), which is clearly below the “nominal design wind conditions” of 8m/s (while standard wind turbines are designed for nominal operation of 10 to 12m/s of wind speed). Because extractable power from wind increases cubically, if everything else is held constant, a factor of above 3, in average power production could be achieved for constant winds around 8 m/s. With the ABM spinning at above 100rpm, this factor could be increased to above 6, leading to, potentially, an average power production around 10kW during the upwards phase.
One very important factor was also evident from testing this prototype: since HAWE is a wind power conversion system that consumes power during part of its operation, it is extremely important to ensure that specifications are met at all subsystems level; otherwise it may happen that one system is jeopardizing the entire system performance and its viability in achieving the objective for which it was designed and manufactured.
15. Proof-of-concept and on-field results – conclusions from testing
An average net mechanical power of 1kW was measured and the results were according to the computation simulations under the same parameters, thus validating the model, the mechanical power could not be translated into electric power due to the extreme inefficiency of the winch.
Despite some testing and grinding issues with the prototype, and all above described system limitations, it was possible to validate the models and therefore provide proof-of-concept with the prototype.
Some issues faced:
a) As explained in section 3.5.3 the ground station efficiency was very low, with the winch (in particular) requiring a very significant pulling force (upwards of 1.5kN) to be able to produce net positive electrical power. Due to normal design safe requirements this component was already over dimensioned by a factor of between 2 to 3. But, because the actual ABM prototype tested was roughly 1/4 of the initially intended size to be tested, the whole ground station ended up being grossly over dimensioned (very likely by a factor of over 10) and this meant that no net positive electrical power cycle was achieved. Calculations thus had to be performed for purely mechanical power considerations, from the load cell measurements taken at the tether cable and swivel point. The electromechanical concept at ground station must, nevertheless, be improved to allow the system to unleash its full potential as well as to optimize its operation parameters;
b) During the upwards (power production) phase, a maximum of only 55rpm could be reached, during the January-February 2014 test campaign. During the same year’s summer test campaign this restriction was overcome and testing with the ABM rotating at 80rpm has been achieved, increasing the Magnus effect speed ratio from 1,5 to around 3,5 (achieved with rotation at 80rpm).
c) The ABM prototype’s net buoyancy is below 50N, which means that, as soon as ABM rotation is stopped the ABM envelope will immediately start to fall, and do so in an unstable fashion compromising the envelope’s structural integrity as well as that of the cable and compromising also the safety of the personnel on the ground station operations; for this reason a rotation to the ABM was imprinted even during recovery, implying much greater effort (i.e. more power) needed to perform recovery;
d) The motors developed for the ABM prototype, by Omnidea, proved a good option with an excellent power-to-weight ratio and the capability to operate above 600V reducing losses through the cable; nevertheless, as this was a task that was later necessary to add, to speed up the development still during the HAWE project, the drivers were not yet designed to also incorporate the function of working as a generator. Therefore the motors could not be used to break the ABM’s rotation; consequently when the ABM was already being recovered at 2 m/s the balloon would sometimes still be rotating at above 30-35rpm, leading to an initial 4kW peak power in the recovery phase. The semi-automatic control system operation during these tests, was not flexible enough to allow changing the synchronicity between balloon rotation and cable speed, minimizing the power consumption due to the balloon’s inertia;
e) Instead of the expected 5-10% energy consumption from the ABM (to keep it rotating), nearly 20% of the average power production was needed. This effect is due to the balloon bending between its “cantilever type” fit in both ends, an un-anticipated effect and is considered, by the HAWE team, a very important experimental result as it is now known that it has to be minimized. Bending can be minimized with higher internal pressure, with bigger rims (structural reinforcement of the tops) or a different rotation system that has been developed throughout this project and tested at sub-system level. In this alternative ABM rotation system, the cylinder is anchored in a third point (through a ring in the middle section) and rotated through this ring. Further info can be obtained in Deliverable 10.4.
16. Economics and financials of HAWE - Assumptions
As defined in previous sections, the HAWE system is constituted by an airborne module (ABM), a ground station a tether cable and a control system that includes communications. In addition, other components such as grid connection, substation and other auxiliary items are also part of the system. The cost structure considers already significant cost reduction when compared with the costs of the prototype, namely the cost of the winch. The main cost drivers are through industrialization of the solution and reduction, raw material and labor required to produce the equipment. It is important to mention that these costs do not reflect adequately the learning curve effect for producing several units. A summary of assumptions can then be provided:
• The main Operation and Maintenance (O&M) activities required are the replacement of the ABM envelope and tether cable, the envelope due to aging and the cable due to cycle fatigue. As a base scenario, it is considered that the cable and part of the envelope is replaced every 2 years. In addition, periodic preventive maintenance shall be carried every 6 months;
• The evaluated project base scenario consists of a “15 units HAWE farm”, each individual unit with up to 250 kW of capacity for a total of 3,75 MW. It was estimated that the farm would require the installation of an aerial electrical line with 5 km of length. It was also assumed that the devices will lose 5% of performance in the first 10 years of operation and 15% up to the 20th year of operation.
• The cost of each component was estimated based on the prototypes’ experience and quotes from supplier and considering that the technology is in a pre-commercial, resulting in a CAPEX for 250kW units of 150KEUR
• As a base scenario, it is considered that the cost of Operation and Maintenance (OPEX) of the farm would be of 5,3%/year of the total investment (CAPEX), which totals approximately 253 k€/year.
• In order to have a comparative analysis against wind turbines a 35% capacity factor was assumed (i.e.: 3.066 equivalent hours of operation at nominal power), which is understood to be very low and not taking advantage of the high altitude winds stability, likely leading to a very conservative approach.
From the economical and financial assumptions the following were based on analysis of the available technical information.
• The Feed-in-Tariff for this project was based in the current Portuguese legislation and as a base case, the tariff is 74 €/MWh.
• Other scenarios, considering experimental feed-in-tariffs, ranging between 162 €/MWh and 208 €/MWh were also modelled in the sensitivity analysis. In either case, the Feed-in-Tariff is limited to 33 GWh/MW or 15 years. After this period, the tariff is the pool price, which is assumed to have an average price of 50 €/MWh.
• In this case, the depreciation was calculated on a straight-line basis, meaning that the capital cost of the equipment was spread evenly over its depreciation period. The depreciation periods were determined according to the prescriptions of Portuguese legislation.
17. Economics and financials of HAWE - Results
With the base assumptions, the project proves to be viable, with a positive NPV and IRR higher than the minimum Return Rate required (WACC). The payback period is 8 years, which is acceptable. The Levelized Cost of Energy (LCOE) is 64,2 €/MWh, which is very competitive with alternative renewable solutions, in fact almost achieving grid parity.
Based on the analysis above, several scenarios were defined:
• Scenario 1 – Base scenario
• Scenario 2 – Increase in energy production by 10%
• Scenario 3 – Reduction in energy production by 10%
• Scenario 4 – Increase in CAPEX by 10%
• Scenario 5 – Reduction of O&M costs by increasing the life time of the cable to 5 years
• Scenario 6 – Reduction of O&M costs by increasing the life time of the cable to 10 years
• Scenario 7 – Increase of the feed-in-tariff to 162 €/MWh
• Scenario 8 – Increase of the feed-in-tariff to 185 €/MWh
• Scenario 9 – Increase of the feed-in-tariff to 208 €/MWh
• Scenario 10 – Reduction of insurance cost
The main results were:
• The HAWE technology is very sensitive to the energy produced. A change of 10% in energy production (scenario 2) allow us to increase the IRR by almost 300 basis points (BPS).
• a decrease of 10 % in energy production (scenario 3) reduces the IRR by 270 BPS, in this situation the project is not profitable, the NPV is negative and the IRR is lower than the minimum required return rate.
• The same conclusion applies if the CAPEX increases by 10% (scenario 4). The NPV of the project is negative and the IRR is lower than the minimum required rate of return.
• The HAWE technology is also very sensitive to the O&M costs (scenario 5 and 6). The base scenario O&M costs are high. If a reduction in O&M costs is achieved, it will have significant impacts in the financial indicators of the project. If a reduction of O&M costs from 5,3% of the total CAPEX to 2,3% is achieved the NPV more than doubles and the IRR increases 350 BPS. The payback time of this scenario is 6 years.
• The HAWE technology is also very sensitive to the feed-in-tariff (scenarios 7, 8 and 9). If the project can obtain an higher feed-in-tariff, the financial indicators improve significantly and the IRR of the scenarios are respectively 37,8%, 46,0% and 54,1%.
• The last scenario (#10), considers the reduction of the insurance premium. Such reduction will allow improving the financial indicators, but it is not as significant and the previous scenarios.
18. Concluding remarks
To conclude, it is very important to for the success of the technology, to implement it with the possibility of extracting the wind energy from high altitude, and therefore achieve high capacity factors which is the highest driving factor to render any project based on HAWE technology as economically attractive as possible. As an example, if there would be a deviation of additional 10% on the base scenario on both capacity factor and CAPEX, then in spite of the negative contribution of the latter (i.e. increase in CAPEX), the overall result would be an improvement in the economic attractiveness (as the increase in capacity factor outweighs the CAPEX effect), meaning that the project CAPEX has a much lower impact proportionally than the capacity factor. Decreasing costs of O&M also has a major impact in improving the return on investment by reducing the payback time; however O&M costs requires a higher technology readiness level (TRL) in order to be properly assessed.
The HAWE team will proceed with the R&D on this technology and with the TRL increase still needed, hopefully until a successful market implementation.
Potential Impact:
Current estimates of the power factor of the system developed in this project, derived from simulations performed and from preliminary results obtained in testing at Ota base, and present in EDPi’s economic feasibility assessment study, indicate that HAWE can be competitive with conventional wind turbines with the distinct advantage of being able to provide electricity on a continuous basis by operating at varying altitudes to capture high altitude wind energy. In the mid to long term, HAWE is expected to meet the sustained demand for energy worldwide as it can operate in a greater range of environments than conventional wind turbines and is easily deployed offshore. The successful development of the HAWE system addresses the global concern for renewable energy, contributing to the SET plan of the EC and furthering Europe´s position as a leader of innovative technology.
Since 2010, when the HAWE project began, the overall wind market has continued to grow, although the 25% Cumulative Annual Growth Rate of the 2000-2009 period could no longer be sustained. After new installations of wind power stabilized in the 2009-2011 period at around 40GW per year, 2012 saw an increase to 45GW followed by a decrease in 2013 to 35GW of new installations. This means that, in 2013, new installations were back to 2008 levels. In terms of worldwide accumulated installed wind power, in June 2014 roughly 336 GW were installed. In the period 2009 to 2013, wind energy finally overtook biomass as the 2nd largest source of renewable electricity and electricity from wind is fast approaching 1000TWh of worldwide electricity generation roughly one third of worldwide electricity generation from hydro. In the same period (2009 to 2013), new turbine installations in Europe have been steady at about 10-11GW per year and Europe grew, in terms of overall wind installations, from 76GW in 2009 to 117GW in 2013.
In 2010 the wind market represented 0,26% of Europe’s GDP with a turnover of around €32,5Bn from direct (€17,6Bn) and indirect (€14,9Bn) employment. Of this turnover, nearly €9Bn were exported, with imports of only €3,6Bn, i.e. in the wind market exports were approximately 2,5 times as large as imports! Employment stood at nearly 240.000 jobs (135.000 direct jobs plus 105.000 indirect jobs). For each €1Bn invested in the wind energy industry, 21.000 jobs were created per full year (source: EWEA). Furthermore, by dividing the amount of created value in the direct employment of turbine manufacturers, component producers and service providers, we get an average of 82.000€ per person (source: EWEA), per year in the wind market, making it standout as a value-added, export oriented market.
Nevertheless, although Europe is still the region of the world with the largest installed wind capacity, China has overtaken Europe in new wind installations and should overtake Europe in accumulated capacity by 2015. In other words, if in 2009, Europe had nearly 50% of all installations worldwide in 2014 these represent less than 35%. Before 2020, with the current trend, Europe will have less than 30% of all wind turbine installations, unless offshore wind installation really “takes off”, after a very gloomy start. In terms of wind electricity production, europe is even further behind China since many installed turbines are very old (i.e. inefficient and small in capacity). Consequently Europe generated “only” 257TWh, less than the amount of electricity from wind generated in China.
The EWEA initially estimated in the 2009 (see the “Pure Power scenario” paper) a total of 230GW of installed wind capacity by 2020, producing 580TWh of energy per year, with 40GW coming from installed offshore wind power. However, new scenarios drafted between EWEA and main sector players in 2014, now point towards 192GW of installed capacity and 442TWh of energy by 2020. Even for this reduced scenario to be achieved, between €90 and €100Bn will have to be invested until 2020 a very considerable investment under the current economic conditions. European turbine producers have approximately 90% of the European market, roughly 30% of US market and less than 10% of the Asian market. European manufacturers Vestas, Enercon, Gamesa and Siemens Wind represent nearly 35% of the whole market share of wind turbine producers but Chinese players are also over 30%.
As anticipated by this consortium, in the proposal submitted in 2009, the maximum size of wind turbines has not changed significantly since 2009. Enercon’s E-127 (currently in 7MW form) remains the largest commercially available turbine. Turbines of 10MW are in development, but these are being tailored for off-shore and will likely only be competitive in “extremely favourable” off-shore wind park environments/scenarios. The evolution of the average turbine size in Europe, in the past 5 years, shows signs of an increase (i.e. total installed capacity divided by the number of installed wind turbines), but this trend does not seem sufficient to fundamentally change the most significant argument, presented by this consortium, in the 2009 proposal: wind turbines face challenges to grow past the 6-8MW size and will be significantly more expensive in the 20MW range, especially if no good offshore locations, to install them, exist!
If conventional wind power has been mostly stagnating in Europe since 2009, as has been described above, the same cannot be said about Airborne Wind Energy (AWE) and HAWE (High Altitude Wind Energy) technologies. It is true that market status has not been achieved so far: currently, no AWE/HAWE is in advance-demonstration phase (i.e. in continuous operation for over 8000 hours, roughly 1 full year) but some industry trends are clear:
• A distinction has emerged between AWE and HAWE;
o AWE seems to have centred around operations without a turbine tower and an operating altitude between 150-300m based on concepts of cross-wind flying kites;
o for HAWE, concepts are less “standardized” (Magnus-effect based concepts coupled together with wind mill ladder type concepts, etc…) and operating altitudes around 400-800m have been experimented with.
• Major topics, in the industry (i.e. both for AWE and HAWE) now focus on:
o R&D metereological assessments, led by Ken Caldeira and Cristina Archer;
o Air traffic implications of AWE/HAWE especially for installation in developed countries where most of the airspace is either blocked or under severe usage restrictions;
o Control of the devices, especially cross-wind kites which will fly at above 50 knots in directions perpendicular to the main wind direction
• Providing demonstrations of capabilities, such as
o Makani Power in 2010
o Altaeros in 2013
o TUDelft (Laddermill Kitepower) in 2012 (700m of altitude operation)
Although it is hard to estimate, over 2000 professionals are currently working in AWE/HAWE concepts of or related work, with over 40 institutions currently in charge of this development. At Omnidea, and as part of this project, approximately 8 jobs were partly (or almost fully) funded during the duration of this project. Overall in the consortium, between 20-25 were almost permanently allocated to this project as per Figure 4 1.
At the current state-of-play, it is wise to introduce here the words of PJ Shepard co-founder of the Airborne Wind Energy Consortium:
“the progress toward working, power-generating, AWE prototypes has been impressive in the last few years. Some companies have proven energy-generation capability, many have demonstrated some level of autonomous control, and a few are planning power farm development”.
But, in the words of the industry: “scaling up past the prototype stage won’t be easy”. Experts in the industry believe that without strong government support, installing even one gigawatt of airborne power could take 20 years or more. Some larger companies such as Honeywell and 3M have shown some interest in the AWE/HAWE technology, but the major traditional wind power developers — Siemens, GE, Vestas, and others — have so far left the airborne designs to the startups.
A recent study by Cristina Archer shows that there are enough areas of the globe where consistent, 10m/s wind conditions exist for at least 15% of the year’s time. From this it is clear that the potential exists for AWE to provide generation power enough for Europe’s needs.
Another study from Colm O’Gairbhith from Loughborough University, published in 2009, compares wind availability at several possible test sites in Ireland and presents a comparison between typical wind speeds at 150-200m and wind speeds at 750m. A rough comparison gives a potential power increase of nearly 101% i.e. roughly double that at lower altitudes for typical wind speeds of around 9,16m/s (at 150m) and 11,86m/s at 750m. These values are based on the equation where wind power, P, increases with the cube of wind velocity, V. The same study points towards a capacity factor of 52% for kites which effectively work in a similar pattern to that of HAWE (a pumping cycle with a reel-in phase). Therefore, assuming that HAWE could also have a 4:1 ratio between the traction phase and the reel-in phase, HAWE could also have a capacity factor of 52% but at the it moment seems to be a little optimistic for HAWE but within the overall capability of the system). This means that there is approximately double the potential energy to extract (101% more) and a roughly 12-20% greater capacity factor (vs. respectively 40% capacity factor for off-shore turbines and 32% for on-shore turbines). Combining both factors it seems realistic to say that power outputs can triple from the use of AWE/HAWE technologies.
According to a report from Carbon Tracking, AWE/HAWE technologies have attracted the attention of the science and engineering communities, with articles having already appeared in the New Scientist, Scientific American (who identified AWE as one of the top five “Greentect” technologies of the future) and the New York Times. The Norwegian state-utility Statkraft, the largest renewable energy company in Europe, has identified Airborne Wind Energy as an energy technology for the future and has invested in technology development. A study, published in 2010 and entitled “Assessing the Potential for an Airborne Wind Energy test-site in Ireland”, presents a short overall description of the “by then” status of technological evolution as well as putting forth 3 potential test sites in Galway and Offaly counties. The test site assessment also presents results from a survey conducted with several identified AWE/HAWE technology developers, such as Makani, Joby energy, Altaeros, etc.. The study provides two budget estimates for site construction: one called “full costs” and another called “start-up” version. Costs vary between 2,5Mn€ (full cost) and 0,9Mn€ (start-up costs). This is very helpful in shedding some light into the benefits of using the Ota test site for free.
In March 2011, Omnidea was contacted by the then Makani power CEO, Corwin Hardham on the possibility of a partnership for Airborne Wind Energy. This possibility was discussed at length by email and a meeting was held at the 2011 Leuven AWEC. Nevertheless, because Makani’s prime interest was in a concept employing wings (which HAWE exchanged in favour of a rotating cylinder concept for the Magnus effect) the potential for future cooperation was judged to be small. On October 23rd 2012, Mr. Corwin Hardham died unexpectedly from cardiac insufficiency; Omnidea’s team hereby honors his memory as one of the AWE community pioneering minds; with Mr. Hardham having himself been the initiator of the talks between Makani and Omnidea, the cooperation between both institutions did not develop any further. In August 2013 Makani was acquired by Google.
During the 4th HAWE Progress Meeting, hosted by EDPI in Lisbon in March 2012, discussions were held with Prof. Uwe Paulsen, from DTU-WE regarding the possibility of exchanging information between the FP7 project DeepWind, coordinated by DTU-WE and HAWE. This possibility is especially interesting regarding the availability of aerodynamic coefficients for water immersed cylinders under a rotation motion, and its importance to the estimation of drag coefficients. Prof. Paulsen has since sent to Omnidea a paper describing the operation of the Deepwind project, as well as its challenges and accomplishments in greater detail. This collaboration is still on-going, especially regarding the estimation and calculation of 3D lift and drag coefficientsfor the Magnus effect.
Even during the proposal elaboration stage it was already clear that HAWE is supported by IPR, in this case this technological development is supported by the PCT patent “Atmospheric Resources Explorer” (PCT/PT2007/000022) which forms the basis for this consortium to undertake work in this Collaborative Project in FP7”. This was already referred to in the proposal and the PCT patent was finally awarded in December 2010, just 3 months after the beginning of the project. The “Atmospheric Resources Explorer” patent is the crucial IPR piece in this technology as it describes both the layout of a typical HAWE system (such as the one tested at Ota, i.e. with an hybrid-LTA ABM, a tether cable and ground station containing a winch) as well as the operation method of HAWE (i.e. through a pumping cycle).
Another patent, obtained at national level and under evaluation at international level, named "Airborne Platform" was submitted, at international level, on the 6th month of this project (i.e. March 2011). This patent describes in detail the ABM concept and design, aiming at protecting industrial property over Omnidea’s currently in development part of the HAWE technology. The invention pertains to aeronautical enginering and consist of an airborne platform that can be built to large sizes without requiring rigid structures of comparable dimensions and which uses both buoyancy and aerodynamic magnus effect for lift.
A third patent, also obtained at national level and under evaluation at international level, named "Submerged Platform" was submitted at international level, on the 19th month of this project i.e. (in July 2012). This patent describes in detail a system under traction stress (such as the mooring platform of a HAWE concept, which means that the ground station can be positioned below the surface, thus not exposed to adverse weather conditions with an exit emerging above the surface to allow tethering of the ABM.
As consequence of the work done by Lankhorst on this project, a patent application is being submitted for the new designed cable. Having as provisional name “Multifunctional Tether”, it will likely still be submitted in the beginning of 2015, likely both as an European/international patent and also in the US. Lankhorst’ push comes from multiple commercial inquiries for an electro-mechanical cable, which, despite will very likely materialize, into a real order, in the near future.No expenses have been presented, by Lankhorst, to the EC, on this matter (i.e. IPR registration). No expenses, relating to this patent, have been (or will be) presented to the European Commission during the HAWE project.
The HAWE website has been online since the initial Kick-Off meeting in Lisbon (Oct. 2010) and a new section for partners was added at the time of the Dubrovnik meeting. The site can be accessed through http://www.omnidea.net/hawe/ and partners (as well as the commission) can access specific contents through the “login” button. The HAWE website was the subject of an update in June 2012 and again in September 2014.
A moodle platform for file sharing was also created in June 2011 but it showed little to no acceptance by the project partners. As a consequence an ftp website, for data and file transfer, was created in September 2011 and has since been used by the partners. Through this ftp website, videos are also available, from the testing activities at Ota. Finally, since 2014, Conferences, articles and publications
Mr. Pedro Silva and Mr. Tiago Pardal attended the Airborne Wind Energy Conference 2011 held in Leuven, Belgium in May 2011. At this conference Mr. Tiago Pardal made a presentation “Omnidea’s system for HAWE-FP7” on behalf of co-authors Mr. Rei Fernandes and Mr. Pedro Silva which is also accessible through the HAWE FTP website.
At the 3rd HAWE meeting, held in September 2011 in Dubrovnik, Croatia, Neven Duic (FSB), Mr. Luka Perkovic (FSB) and Mr. Pedro Silva (Omnidea) presented a joint FSB-Omnidea paper on the CFD simulations performed during the HAWE project, entitled “Harvesting High Altitude Wind Energy for Energy production: a feasibility study of Magnus effect concept”, at the SDEWES Conference, on Sustainable Development on Energy, Water and Environment Sytems.
Also in September 2011 at the same conference, Prof. Josko Deur submitted an article related to the energy storage solutions possible to implement for an application such as HAWE in which a relatively high number of storage cycles for relatively low energy requirements but with relatively high discharging power requirements, are compared for the virtues of purely mechanical storage (i.e. a flywheel based solution) versus an electric solution employing supercapacitors.
In June 2012, Prof. Tatjana Haramina (from FSB University Department) presented a paper at the “Energy and Materials Research Conference” held in Malaga, Spain entitled “Permeation of hydrogen and other types of buoyant gases through foils for balloons and their tensile properties” from her work on diffusion properties of the inner film layers performed for HAWE together with Mr. Pedro Lima (Omnidea), mentioned as co-author.
An article has been published by N. Duic, R. Penedo and L. Perkovic for the “Energy journal” (Elsevier) titled "Estimating the spatial distribution of high altitude wind energy potential in Southeast Europe". It focuses on the development of a tool to find available spots to deploy an High Altitude Wind Generator based on the population density, the existence of roads and railroads, and other significant obstructions to the use of high altitude devices. For this study HAWE was used as the base reference for the assessment in the Southeast Europe region.
HAWE’s most important major dissemination activity was the participation in the Springer Verlag book “Airborne Wind Energy” published in 2013 by the editor Professors, Roland Schmehl (TU Delft), Uwe Ahrens (Hochschile Heilbronn, NTS Energy systems) and Moritz Diehl (University Leuven) to be part of the first comprehensive publication on airborne and high altitude wind energy. The HAWE concept makes up the 29th chapter of the mentioned book as a case study. For this publication, detailed HAWE simulations were produced, along with a detailed concept comparison with typical kite-based solutions. As a consequence of the above publication HAWE was also present at the 2013 AWEC (Airborne Wind energy conference), held in September 2013 in Berlin, through both a poster presentation and an oral presentation by Mr. Tiago Pardal, Mr. Rei Fernandes and Mr. Pedro Silva.
A final dissemination activity was undertaken in October 2014 when the Portuguese TV channel SIC (http://sic.sapo.pt/) undertook an in-depth news article about HAWE, broadcast on 14.10.2014 at the main SIC channel (on the evening news) and which can be retrieved at the following link (http://sicnoticias.sapo.pt/programas/futurohoje/2014-10-15-Usar-o-vento-para-produzir-energia-mais-barata-). This news item is part of the program “Futuro Hoje” (the future today) of which Omnidea’s technology represents one chapter (this program features, on multiple occasions, green energy technologies). This article was produced during a testing day at Ota Air Force base.
List of Websites:
http://www.omnidea.net/hawe/
Contact details:
Nuno Fernandes
Omnidea, lda
Room 2.1 - Ed. VIII
Campus da FCT
2829-516 Caparica, Portugal