GB2630911A - Large vertical axis wind turbine - Google Patents

Abstract

A large vertical axis wind turbine (VAWT) having a hub 4 comprising an annular drive track (fig.3, 12) cooperating with drive rollers 16 (that may be tapered) that support the weight of the rotor wherein friction between the rotating drive track and the drive rollers turns the drive roller that turn an axle 17 to drive a generator 18. There may be a gearing system. The hub also comprises vertical axis bearings for supported the tower spaced around the annular track. The hub may comprise a braking system with brake pads 23 for use in emergency situations. The turbine blades (fig.1, 1) may be assembled by joining two aerofoil sections together at their root ends to make a double length combination blade. The wind turbine may also comprise a parachute that is triggered during overspeed conditions that are detected by a rotor overspeed sensor.

Description

Large vertical axis wind turbine This invention relates to vertical axis wind turbines that have a rotor swept area much larger than horizontal axis wind turbines with blades of similar length, with means to convert the consequent high torque shaft power output to electricity efficiently and using proven technology as well as means to react the large downwind force on the rotor and multiple means to ensure that the rotor can be safely brought to rest as and when needed, together leading to substantial reductions in the cost of energy from offshore wind farms and enabling their cost-competitive use in deeper waters and further offshore.

The efficient modern vertical axis wind turbine (VAWT) was invented in France in 1925 by Darrieus but attracted little interest until it was rediscovered by South and Rangi at the National Research Centre in Ottawa, Canada, in the late 1960s. With interest in all renewable sources of power stimulated by the 1973 oil crisis it -as well as the more usual horizontal axis wind turbine (HAWT) -then became the subject of considerable research and development activity.

Early studies post-1973 in the USA, Canada and elsewhere focussed mostly on VAWTs with the curved blades and 'egg-beater' configuration proposed by Darrieus. In particular Sandia Laboratories in the USA were funded by the US Department of Energy through the 1970s and subsequently to thoroughly investigate the potential of VAWTs, and their well-documented activities included the construction and test of a two-bladed, 34 m diameter, 500 kW rated, 'eggbeater' configuration test-bed VAWT, which commenced operation in 1988. In the UK, based on the work done by Musgrove at Reading University in the mid-1970s, the main focus for VAWT R & D was the straight-bladed H-configuration, culminating in the construction and testing of a two-bladed, 35 m diameter, 500 kW rated machine in 1990.

However by about 1990 it was evident that VAWTs -whatever the configuration -were somewhat less efficient than contemporary HAWTs. VAWTs were consequently deemed to offer no competitive advantage over the HAWTs that were already in widespread use, and the commercial development of wind power has since been dominated by 3-bladed HAWTs with blades upwind of the tower, the same configuration as was used for the first generation of successful grid-connected wind turbines developed in Denmark in the late 1970s.

Through the 1980s it became clear that larger wind turbines delivered lower cost electricity, and the 15 m diameter, 55 kW rated machines of the early 1980s were succeeded by progressively larger wind turbines. The enormous potential of offshore wind farms began to be developed in the early 2000s using wind turbines that by then had a rotor diameter of about 90 m and a power rating of about 2 MW. Constructing wind farms offshore requires balance of plant costs (foundations, cabling, installation etc. ) that are much higher than on land and for a given overall wind farm capacity -using a smaller number of larger wind turbines gives substantial cost savings. Hence a continuing drive by manufacturers to develop progressively larger wind turbines, with the result that by 2021 the turbines being deployed offshore had -on average -a power rating of about 7.5 MW and a diameter of about 160 m, and by 2025 the rated power is expected to average 13 MW with a rotor diameter of about 220 m, but even at this size the turbine cost will be little more than one third of the total offshore wind farm cost.

Though the commercial development of wind power has been dominated over the past four decades by the three-bladed horizontal axis wind turbine, R & D on vertical axis wind turbines has continued, most notably at the Sandia Laboratories in the USA, where -funded by the Department of Energy -their investigations have continued through to the present. Their 2012 report 'A retrospective of VAWT technology' by Sutherland, Berg and Ashwell, SAND2012-0304, summarises not only their own work but the work done in recent decades on large VAWTs by others. The 2019 report 'A historical review of vertical axis wind turbines rated 100 kW and above' by Mollerstrom, Gipe, Beurskens and Ottermo, Renewable and Sustainable Energy Reviews 105 (2019) pp.1-13 gives further information on medium-sized and large VAWTs that have been constructed. And the 2019 review by Kumar et al. 'Review on the Evolution of Darrieus Vertical Axis Wind Turbine: Large Wind Turbines' Clean Technol. (2019), 1, 205-223 also covers some interesting large VAWT concepts that have been proposed in recent years, but which never progressed to construction. And most recently the Swedish company SeaTwirl has commenced construction of their 1 MW rated S2 offshore floating VAWT. However with just the one exception noted below none of the large VAWTs that have been built or proposed in recent decades has any significance in relation to the features that characterise the present invention. The sole exception referred to above is the 'L-180 Poseidon' Darrieus wind turbine proposed by Ljungstrom in 1980, though his concept was fundamentally flawed for reasons given subsequent to the detailed description of the present invention.

The pressure on wind turbine manufacturers to deliver low-cost electricity in offshore waters that are deeper and further offshore has sustained the drive to develop larger wind turbines, and prototypes with rotor diameters up to 260 m and power ratings up to about 18 MW have now been built and are under test. But turbines this size require components such as the fibre-composite blades, as well as bearings, castings and gearbox/generators, that are pushing the limits of present technology. Further incremental increases in the size of HAWTs are to be expected in due course but the VAWT that is the subject of this invention uses existing blade technology to more than double the rotor swept area, and hence the corresponding power output, with consequent substantial economic benefits.

According to this invention a substantially vertical double length combination blade is made by joining together by their root ends two blades (each comparable with the blades already used by large modern HAWTs); this double-length combination blade is then attached to the hub of a VAWT rotor via a substantially horizontal cross-arm of similar length, with the option of attaching additional combination blades via additional cross-arms to the same hub. This VAWT configuration approximately doubles the rotor diameter -and more than doubles the rotor swept area -by comparison with a HAWT using the same length blades. However this VAWT then has an optimum rotational speed that is substantially less than half that of the corresponding HAWT, resulting in a VAWT rotor torque that is substantially more than four times higher than the HAWT rotor torque. This invention describes how to convert the VAWT's very high torque power output efficiently to electricity whilst also describing how to support the rotor weight and how to react the very large downwind force on the rotor. The invention also describes the means to address what has -until now -been a major problem for all large VAWTs, how to bring them safely to rest in emergency conditions.

The invention is further described with reference to the accompanying drawings in which: Figure 1 shows the general arrangement of the wind turbine Figure 2 shows a blade/blade/cross-arm joining structure Figure 3 shows the rotor hub, viewed from underneath Figure 4 shows a side view of the rotor hub Figure 5 shows a side view of the tower top Figure 6 shows a plan view of the tower top Figure 7 shows an additional side view of the tower top Figure 8 shows a particular example of a cable-stayed 55 MW VAWT Figure 9 shows a blade (or outer cross-arm) cross-section In figure 1 two aerofoil cross-section blades, 1 and 2, each comparable in length and method of construction to those used for large HAWTs, are joined together by their root ends to make a double-length combination blade which is attached to the outer end of a cross-arm 3 whose inner end is attached to the flat disc-shaped hub 4. The figure shows an H-configuration rotor with two cross-arms -each with its attached combination blade -but rotors with one, three or more cross-arms are options that may be preferred for some applications. The rotor turns about a central vertical axis and its power output is converted to electricity within the hub by means that are described below. The rotor is supported by the tower 5 and the foundation structure 6. Figure 1 shows the wind turbine in an offshore location with the sea surface 7 and the sea bed 8, but the several advantages of the invention are such that it may also be preferred for many on-land applications. Though the aerofoil cross-section of a VAWT differs from that of a HAWT the peak loads on each upper blade 1 and each lower blade 2 are of similar magnitude to the peak loads on the similar length blades of a HAW'T, and the method and materials of construction are also similar with the main loads taken by carbon fibre composite and/or glass fibre composite spar caps, separated by carbon fibre composite and/or glass fibre composite shear webs. Aerofoil profiles for VAWTs have been the subject of much research and development in recent years; benefiting from this the profile used would typically be the DU12W262 (or similar).

The root ends of each upper blade 1 and each lower blade 2 may be adapted to facilitate their being bonded or otherwise joined together and then -as a combination blade -being joined to the outer end of each cross-arm 3; alternatively, and as shown in figure 2, a short blade/blade/cross-arm joining structure 9 allows the use of blades with standard root-end fixing arrangements. This blade/blade/cross-arm joining structure could either be fabricated from carbon fibre composite and/or glass fibre composite, or it could be made -like the hubs of many HAWTs -from ductile iron. In either case the blade/blade/ cross-arm joining structure 9, with the upper blade 1 and the lower blade 2 attached, would then be bolted (or otherwise joined) to the outer end of the cross-arm 3.

As noted above the inner end of each cross-arm is attached to the flat disc-shaped hub, which is shown in figures 3 and 4. Though the figures show a hub with two attached cross-arms for some applications rotors with one, three or more cross-arms -each with its attached combination blade at the outer end may be preferred. The cross-arm 3 that attaches each combination blade to the disc-shaped hub needs to be well streamlined, so as to minimise the power loss that results from the cross-arm drag. It could be made from carbon fibre composite and/or glass fibre composite but the former is very costly and the latter is very flexible, which would lead to large deflections, so a hybrid construction is preferred with an inboard length that typically uses steel -for stiffness and lower cost -and an outboard length made using carbon fibre composite and/or glass fibre composite. The use of steel for a substantial part of the cross-arm makes it much heavier but the consequent high bending moment where the cross-arm 3 joins the hub 4 -can be greatly reduced by using a cable stay 30 extending from part-way along the cross-arm 31 to a pylon 32 fixed on top of the rotor hub, as is indicated in figure 8.

Figure 3 views the rotor hub from its underside and figure 4 shows a side view through the section A-A indicated in figure 3. The primary hub structure is an annular box beam 10 with a rectangular cross section, or its structural equivalent, with a superstructure 11 to provide weather protection for the electrical generator and other equipment that is fixed to the tower top as well as to provide -if required -additional strength to the hub. For strength and stiffness the primary hub structure is made from steel. On the underside of the hub a flat annular circular track -the drive track, 12 -is attached, with its centre on the rotor axis, and through this the weight of the rotor as well as the torque it produces is transferred to the tower. A second track -the cylindrical reaction track, 13 -is also attached to the hub with its centre also on the rotor axis and it is through this reaction track that the large downwind force on the rotor is reacted. Both the drive track and the reaction track are made from a rigid material with good wear-resistant properties such as steel, though other materials with similar properties could be used instead. Both tracks are preferably made using a number of close-fitting replaceable segments so that if need be they can be individually replaced.

Figure 5 shows a side view of the tower top where all the machinery needed to convert the rotor's power output to electricity at the required voltage and frequency, as well as all the equipment needed to control the rotor, is located in a generating room whose floor 14 is fixed to the tower top, and whose rotating wall and roof are provided primarily by the rotor's reaction track 13 and superstructure 11. The dashed lines show how the rotor hub sits in position on the tower top. For weather protection a non-structural cylindrical skirt 22 extends down from the outer cylindrical surface of the annular box beam 10 to the level of the generating room floor 14, leaving just a small gap between the bottom of the skirt and the circular perimeter of this floor.

The rotor weight is supported on three or more cylindrical rollers -the drive rollers, 15 -which are radially aligned and substantially horizontal and positioned directly underneath the drive track 12 and distributed evenly around it. These rollers, with their axle bearing housings 16, are comparable with those used in steel rolling mills and so also need to be made from steel or some comparable strong and inflexible material. Figure 6 shows a plan view of the tower top arrangement when just three drive rollers 15 are used. The side view shown in figure 5 is a view through the section B-B that is indicated in figure 6 (though for clarity other equipment that would be visible in the background of this section BB has been omitted). Though the drive rollers 15 could be fixed via their axle bearing housings 16 directly to the floor 14 of the generating room there is the option of fixing them to a platform 19 which is attached to the floor using short stroke hydraulic actuators 20, so that each drive roller can be raised or lowered by a short distance.

As the rotor turns friction between the moving circular drive track 12 and the drive rollers 15 make the latter rotate. Provided slip is avoided the surface speed of the drive rollers is equal to the circumferential speed of the drive track; and since the radius of the drive track is very much larger than the roller radius this means that the rollers' rotational speed is very much greater than that of the rotor. Slip between the drive roller and the drive track can be avoided by appropriate choice of the roller diameter, the weight it supports and the friction coefficient between the track and the roller. This friction contact between the drive track and the drive rollers transfers the low speed, high torque, power produced by the rotor to a much faster and much lower torque power output available from the axle of each of the drive rollers. A shaft extension 17 from the axle of each drive roller transfers this power to the electrical generator 18, either directly or via a gearbox, and the relatively high speed and relatively low torque mean that both these can be of conventional design and well within the limits of existing equipment. Electrical cables (not shown) from each generator then transfer the power to the power conversion and control equipment which may for convenience be located in an inner room 21 that can be fully weather protected and -if required -be temperature controlled. Whether the generator is driven directly from the output shaft or via a gearbox there is benefit in allowing the rotor speed to vary slightly as it turns, so that the rotor inertia can be used to help smooth the rotor torque.

The electrical generator 18 (as well as its gearbox, if this option is chosen) could be fixed directly to the generating room floor. However its (their) weight is just a fraction of the weight supported by each drive roller so if the option of fixing the drive roller and its axle bearing housing to an intermediate platform 19 is chosen then the generator 18 (with its gearbox, if used) can be fixed to the same platform; then each complete power train -from drive roller 15 through to its generator 18 -can be raised or lowered as a unit by means of the short stroke hydraulic actuators 20 that attach the platform 19 to the floor. (Note that attaching each platform 19 to the generating room floor needs to be done in such a way that the large horizontal force -tangential to the drive track -that is transferred by the drive track to each drive roller and its axle bearing housing is also reacted. This could be achieved by using substantially horizontal swinging links each attached at one end to the platform 19 and with the other end attached -directly or indirectly -to the generating room floor 14. Alternatively the hydraulic actuators 20 could have their axes inclined to the vertical so that in addition to supporting the rotor weight they also react this horizontal force).

The benefit of having the power train from each drive roller 15 to its associated generator 18 mounted on a common platform 19, that can be raised or lowered hydraulically, is that it enables a fail-safe braking system. Figure 5 shows a large brake pad 23 on its support structure 24 which is fixed to the generating room floor with the flat upper surface of the brake pad positioned directly underneath and just a short distance below the drive track 12. Several such brake pads are distributed around the floor of the generating room, and directly underneath the drive track, though figure 6 shows just three. When emergency braking is required the hydraulic power system that powers the actuators supporting each platform can then be depressurised, allowing the rotor to drop under its own weight the short distance needed to bring the drive track down onto the emergency brake pads; with the full weight of the rotor then taken by the brake pads the rotor can be safely brought to rest and kept safely stationary, even if there is a total loss of power within the generating room.

A simple aerodynamic alternative means of providing emergency braking is to use drag parachutes. Drag parachutes have been widely used on military aircraft for many years, as an aid to safe landings. In more recent years they have become a common feature on many light aircraft for use in an emergency (such as pilot incapacity) so that they can bring the whole aircraft safely down to the ground. For light aircraft use these parachutes -known as rescue parachutes are commercially available with deployed canopy areas up to about 300 square metres and they can be deployed at aircraft speeds up to about 100 m/s. Quite commonly they will be positioned immediately behind the cabin of the light aircraft, and when activated their rocket-assisted deployment is through a frangible hatch in the roof just aft of the cabin.

On the VAWT a similar parachute system used for emergency braking would be positioned near the outboard end of each cross-arm and could be installed just aft of the primary load-bearing structure formed by the two spar caps 35 and two shear webs 36 shown in figure 9. When activated in an emergency, preferably by an autonomous over-speed sensor positioned close to the parachute container, the parachute's rocket-assisted deployment would then be through a frangible hatch designed into the cross-arm's lightweight rear structure. One appropriately-sized parachute per cross-arm would be sufficient to slow the rotor down to a safe low speed, though given the simplicity and relatively low cost of this emergency braking system a further level of protection could be provided by having a second wholly independent drag parachute emergency braking system installed within each cross-arm. Access to the parachute containers for inspection and -if ever need be -for replacement would be through a walkway within the cross-arm.

The circumferential speed of the drive track 12 increases with the distance from the rotor axis, so the speed of the track where it is in contact with the outer end of the drive roller 15 is slightly faster than at the roller's inner end. To avoid slipping or skidding at any point along the roller its outer end needs to have a slightly larger diameter than its inner end, however the necessary roller taper angle is only about 1 degree. Since the surface of the drive track is horizontal this taper means that the inner end of the roller axis is slightly higher than its outer end, with the result that the axis of the drive roller is at a small angle -about 1 degree -to the horizontal, but this is not a problem. (The drive shaft of HAVVTs is usually inclined by about 5 degrees to the horizontal so as to enhance the clearance between the blades and the tower). If desired the drive track could be made so that its flat surface is at a small angle to the horizontal, such that the axis of each drive roller is then horizontal. (The emergency brake pads 23 would then also need to be made such that their top surface was at the same small angle.) The primary braking system, used for all normal operational purposes, is provided by having several spring-applied/hydraulically-released brakes distributed around the periphery of the generating room, just inside the cylindrical reaction track, with brake pads shaped to match the curvature of this track. One of these primary operating brakes (with its support structure 25) is shown in figure 5, fixed to the generating room floor and with its brake pad 26 adjacent to the reaction track 13. Figure 6 shows three of these primary brakes distributed around the inside of the reaction track but the illustration is not to scale and there is room for more should this be required. And to minimise wear on the reaction brake pads the initial deceleration of the rotor -when required would normally use the electrical generators to provide regenerative braking. If required the primary braking system can also be used to bring the rotor to a complete stop from full speed, as the heat capacity of the reaction track is such that this would only raise its temperature by a few degrees.

The downwind force on the VAWT rotor is large, substantially horizontal and acts at approximately hub height. It is reacted using the cylindrical reaction rollers 27 (whose axes are vertical) which are fixed via their axle bearing housings 28 and support structures 29 to the generating room floor 14, as is shown in figure 7, which is a side view of the tower top through the section C-C that is indicated in figure 6. For clarity other equipment that would be visible in the background of this section has been omitted; and the dashed lines show how the rotor hub sits in position on the tower top. Figure 7 shows two of the reaction rollers that are needed (with their support structures) and the plan view of the top of the tower, figure 6, shows six reaction rollers (with their support structures) distributed around the periphery of the generating room and positioned such that the reaction rollers are in contact with the reaction track 13 and able -therefore to react the downwind force on the rotor, regardless of the wind direction. The rotor's large downwind force is reacted predominantly by those reaction rollers that are on the upwind side of the generating room; this force is of similar magnitude to the share of the rotor weight that is supported by each of the drive rollers; the reaction rollers 27, with their axle bearing housings 28, are therefore also comparable with those used in steel rolling mills and therefore also need to be made from steel or some comparable strong and wear-resistant material. For very large rotors more than six reaction rollers would be required.

As so far described the VAWT that is the subject of this invention has a rotor swept area more than double that of a HAWT using the same length blades, together with means to convert the low speed, high torque, rotor power output to a relatively high speed, low torque output that can be efficiently converted to electricity within the rotor hub, plus means to support the rotor weight and to react the downwind force on the rotor, plus several independent means to bring the rotor safely to rest, all using proven technology. However this invention also covers optional additional features, described below.

Airflow drawdown The one or more blade-supporting cross-arms (3, in figure 1) of a VAWT according to this invention must -as previously noted -be carefully streamlined so as to minimise drag and the consequential reduction in the rotor output power. This could be achieved by giving the cross-arm a NACA 0025 (or equivalent) symmetrical aerofoil section, however there is considerable benefit in using a modern asymmetric aerofoil section in the DTU FFA-W3 series (or equivalent). With the cross-arm's aerofoil section chord line set parallel to the horizontal the FFA-W3-270 aerofoil, for example, gives a section lift coefficient of approximately 0.5 with a section drag coefficient close to 0.01; and the lift coefficient can be substantially increased -with minimal effect on the drag -by having the cross-arm's chord line set at a small positive angle to the horizontal. Then as the rotor turns the airflow over each cross-arm produces lift. (Somewhat more than average when the combination blade at each cross-arm's outer end is moving parallel to and into the wind, somewhat less when the blade is moving downwind, and at a near average level when the combination blade at each cross-arm's outer end is moving across the wind.) This cross-arm lift usefully reduces the overall downforce on each cross-arm due to its weight (and the weight of its attached combination blade). But much more importantly the upward vertical force on the one or more cross-arms results in an equal but opposite downforce on the airflow through the rotor. With a lift coefficient on each cross-arm close to unity this drawdown of the airflow approaching the rotor can be substantial, with the flow through the rotor inclined down by close to 20 degrees relative to the horizontal. As the wind speed above the ground increases with height this means that the airflow approaching the upwind-crossing blade has a higher speed than would be the case with no airflow drawdown, so increasing the power output. And the flow inclination also means that the downwind-crossing blade experiences an airflow largely undisturbed by the upwind blade's interaction with the wind, giving a further increase to the rotor power output.

The airflow pulled down into the rotor from ahead and above necessarily displaces air flowing below the rotor, promoting mixing in the rotor wake up to a height well above the rotor, and this enhanced mixing is to the benefit of downstream wind turbines.

Free rollers Each free roller is cylindrical, radially aligned and substantially horizontal, and positioned directly underneath the drive track 12 and adjacent to one of the drive rollers. Like the drive rollers 15 each free roller is attached via an axle bearing housing 16 to a platform 19, which in turn is attached to the generating room floor using short stroke hydraulic actuators 20 so that each free roller can be raised or lowered through a short distance. However free rollers, unlike drive rollers, have no power output shaft and when in use can turn freely with minimal friction. Their primary use is to enhance the part-load efficiency of the wind turbine by enabling one or more of the drive rollers to be off-loaded. All wind turbines spend much of the time operating at power levels well below their rated maximum power. Under these conditions a free roller can be raised so that the free roller replaces the drive roller in helping to support the rotor weight; power losses associated with operating the drive roller's electrical generator (and gearbox, if used) are then avoided. As the load supported by each free roller (when in use) is comparable with the load usually supported by the adjacent drive roller the free rollers (and their axle bearing housings) also need to be made from steel, or some comparable strong and inflexible material.

A secondary benefit of having free rollers is that their use facilitates inspection and/or maintenance activities on the adjacent drive rollers and their attached electrical generators (and gearboxes, if used) without requiring that the wind turbine be shut down.

FsLslyei;xiLadij12rrn dump. Any interruption to a wind turbine's grid connection necessitates rapid shutdown, otherwise the rotor's power output can only be absorbed by an increasing rotor speed which would soon result in catastrophic failure. Often the interruption to the grid connection is short lived, requiring that the wind turbine be restarted almost immediately. With a VAWT according to this invention there is ample room in or just below the generating room to accommodate a large liquid-filled tank. Electrically resistive elements immersed in this liquid provide a means to absorb the rotor's power output for many tens of seconds and if the grid connection is restored within this time the wind turbine can immediately re-commence exporting power to the grid. A seawater filled tank just 1.5 m high and 4 m diameter has the capacity to absorb the power output of a 55 MW wind turbine for well in excess of a minute.

According to this invention a particular example of a large VAWT is shown in figure 8, which depicts a two-bladed turbine with a rotor diameter of 560 m and with a rated power output of 55 MW at a wind speed of just under 12 m/s. The combination blade attached to the outer end of each cross-arm has an overall length of 246 m and comprises the upper blade 1 and the lower blade 2, each 120 m long, with the blade/blade/cross-arm joining arrangement adding a further 6 m to the combination blade length. The rotor swept area of 138,000 square metres is three times larger than the swept area of a HAWT using similar 120 m long blades. Based on the well documented characteristics of the Sandia 34 m diameter 500 kW VAWT test bed the 120 m long blades of the 55 MW VAWT described here have a constant chord of 15 m. And using the DU12W262 aerofoil profile developed specifically in recent years to improve the performance of VAWTs the blade thickness is 3.9 m. As with modern HAWTs the blades are made with mostly carbon fibre composite spar caps 35 (see figure 9) joined by mostly carbon fibre composite shear webs 36. The nose section 37 and the tail section 38 are relatively lightly loaded and made using mostly glass fibre composite. The approximately 6 m width of the spar caps varies in thickness from approximately 50 mm at the root of each blade (adjacent to the blade/blade/joining structure) down to near zero at the blade tip. The overall blade mass is then approximately 80 tonnes. The peripheral speed of the combination blade is 60 m/s at the 55 MW rated maximum power output; this corresponds to a rotor speed of just 2.1 rev/ min, and an average rotor output torque of 257 MNm.

The 120 m long upper blade and the 120 m long lower blade are both fabricated with 3 m long extensions to their root-end spar caps and shear webs, and these are bonded and bolted together with double cover butt joints, using cover panels that are also mostly made using carbon fibre composite. The mostly carbon fibre composite spar caps and shear webs within the outer cross-arm 33 are made with increased thickness at their outboard ends so that the 246 m long combination blade can then be bolted on in a similar way to that used to attach the blades of multi-megawatt HAWTs to their hubs. The joining structural arrangement described above is enclosed within a mostly glass fibre composite structure that maintains the aerofoil cross-section of the blade over this central region and which provides a well-faired transition to the aerofoil cross-section of the outer cross-arm.

The outer part 33 of each cross-arm is approximately 120 m long, similar to the length of both the upper blade and the lower blade and with a similar 15 m chord, and is of similar construction to these blades with mostly carbon fibre composite spar caps and shear webs and mostly glass fibre composite nose and tail sections. To achieve as large an airflow drawdown as is reasonably practicable the outer cross-arm has a DTU FFA-W3-270 aerofoil section and has its chord line pitched up by about 2 degrees relative to the horizontal. With this aerofoil section and a blade chord of 15 m the outer cross-arm thickness is 4.0 m. The spar cap thickness needs to be about 20 mm at its outer end, increasing to about 40 mm at its inner end (where it joins the inner cross-arm 34) and its mass is approximately 110 tonnes. And there is benefit if the outer cross-arm is made with a pre-curve so that when the rotor is in operation, and loaded by its weight plus the weight of the attached combination blade (only partially alleviated by cross-arm lift) its outer end is approximately level with the reaction rollers 27 in the hub.

Carbon fibre composite construction is expensive and -by comparison with steel -relatively flexible so the inner part of the cross-arm 34 is made mostly from steel. To enhance its stiffness and strength the inner cross-arm 34 is made with the thicker DTU FFA-W3-360 profile, which with the chord unchanged at 15 m gives it a maximum thickness of 5.4 m. And since the inner cross-arm is thicker than the outer cross-arm the inner end of the outer cross-arm is gradually flared out (over a length of about 10 m) to match the profile of the inner cross-arm to which it is joined with a bolted lap joint (or equivalent). The primary load carrying structure within the inner cross-arm is a relatively thin-walled steel box beam that is 6 m wide (to match the spar caps on the outer cross-arm) and up to 5.4 m deep, with the upper and lower surfaces curved to match the required aerofoil cross-section; the lightly loaded nose and tail sections can -as with the blades and outer cross-arm -be made using glass fibre composite.

The use of a cable stay 30 that runs from the top of the 80 m high, steel, multi-tubular pylon 32 (fixed to the top of the 50 m diameter disc-shaped hub 4) to a fixing 31, located close to where the outer cross-arm 33 is joined to the inner cross-arm 34, greatly reduces the bending stresses in the inner cross-arm; an average inner cross-arm box beam wall thickness of approximately 50 mm is then sufficient and this gives an inner cross-arm mass -mostly steel -of approximately 1100 tonnes. The cable stay 30 typically comprises three 200 mm diameter galvanised steel wire ropes positioned in-line, chord-wise, so as to keep their drag to a minimum. The inner cross-arm 34 is bolted on to the 50 m diameter hub 4 whose primary load-carrying structure is a 6 m by 6 m annular box beam 10 (as shown in figures 3 and 4) made from steel and with a wall thickness of approximately 30 mm. Attached to the hub underside is the annular steel drive track 12, and the cylindrical steel reaction track 13 is attached to the inside of the hub's annular box beam. The overall mass of the complete rotor, including its hub and pylon, is then approximately 4200 tonnes, over 90% of which is steel.

The rotor weight of about 41 MN less the approximately 4 MN aerodynamic lift on the cross-arms is supported by three steel cylindrical drive rollers 15, each about 3 m long, substantially horizontal and radially aligned and with an average diameter of about 0.75 m, and comparable with the steel rollers widely used in steel rolling mills. As previously noted the drive rollers are slightly tapered, with their outer ends slightly larger than their inner ends, so as to avoid slip between the drive track and the drive roller surface at any point along the roller length. Each drive roller with its axle bearing housing 16 is fixed to an intermediate platform 19 which is attached to the generating room floor using short stroke hydraulic actuators 20, so that each drive roller can be raised or lowered through a short distance. As the VAWT rotor turns the drive track turns with it and friction between this steel track and the steel drive rollers makes the latter rotate; each drive roller's one-third share of the total rotor power output is then delivered via its axle extension shaft 17 to the electrical generator 18, either directly or through a gearbox. Friction ensures that the circumferential speed of the roller equals the circumferential speed of the drive track where they are in contact; since the roller radius is close to 0.375 m and the drive track radius is approximately 23 m the drive roller's rotational speed is about 61 times faster than the rotor speed. Each drive roller's nearly 20 MW one-third share of the rotor's peak power output is therefore delivered via its axle extension shaft 17 to its electrical generator 18 (either directly or via a gearbox) at a shaft speed of about 125 rev/min; the corresponding shaft torque of just 1.4 MNm is small by comparison with multi-megawatt modern HAWTs.

As is indicated in figure 5 the electrical generator 18 (plus gearbox, if used) is fixed to the same intermediate platform 19 as the drive roller 15 and its axle bearing housing 16. The short stroke hydraulic actuators that are positioned between each platform 19 and the generating room floor 14 therefore support each drive roller's share of the rotor weight (less its share of the aerodynamic lift on the cross-arms) plus the weight of the drive roller and its axle bearing housing and axle extension shaft, as well as the electrical generator (and gearbox, if used). Then when required, and in particular in an emergency, the hydraulic system used to power the hydraulic actuators is depressurised so that all three drive rollers -and the rotor they are collectively supporting -can drop down through the short distance onto the emergency brake pads 23 (distributed around the periphery of the generating room floor and positioned directly underneath the drive track) and the rotor is then safely brought to rest. Note that the large horizontal tome (tangential to the drive track) that results from friction between the drive track and the drive rollers (and through which the rotor power output is transferred to the drive rollers) must also be reacted by the short stroke hydraulic actuators 20, which can readily be achieved by inclining their axes to the vertical (by about 15 degrees).

The large and substantially horizontal downwind force on the rotor is reacted using six cylindrical steel rollers each approximately 3 m long and with a diameter of about 0.75 m fixed -with their axes vertical -via their axle bearing housings 28 and support structures 29 to the generating room floor 14, as shown in figure 7. And as is shown in figure 6 these reaction rollers are distributed around the periphery of the generating room and positioned such that the rollers are in contact with the reaction track 13 and able -therefore -to react the downwind force on the rotor regardless of the wind direction.

The reaction track is also used by the primary braking system. This utilises three or more spring-applied/hydraulically-released brakes distributed around the periphery of the generating room, just inside the cylindrical reaction track, with brake pads shaped to match the curvature of this track. One of these primary operating brakes is shown 25 in figure 5, fixed to the generating room floor and with its brake pad 26 adjacent to the reaction track 13. Figure 6 shows three of these primary brakes distributed around the inside of the reaction track but there is room for more if required. Note that this primary braking system will usually be deployed for normal operational purposes when the rotor is supported by the drive rollers but it continues to contribute effective braking even when the rotor has dropped down onto the emergency brake pads. And operational braking would usually be augmented by using the electrical generators to provide regenerative braking.

As described this 560 m diameter, 55 MW VAWT has two independent braking systems, both able to bring the rotor to rest in an emergency. If deemed necessary an additional aerodynamic emergency braking system can be provided by installing a drag parachute about halfway along each outer cross-arm 33, positioned within the cross-arm's tail section, and just aft of the rear shear web 36. Each parachute's rocket-assisted deployment (initiated by an autonomous over-speed sensor positioned close to the parachute container) would be through a frangible hatch built in to the tail section's glass fibre composite structure.

The three nearly 20 MW electrical generators powered by the output shafts of the three drive rollers may be either direct drive (facilitated by the 125 rev/min shaft speed) or of more conventional design, with a speed-increasing gearbox intermediate between each drive roller and generator. Whichever option is chosen the generators should allow the rotor speed to vary by a small percentage within each revolution so that the high rotor inertia can be used to greatly reduce the torque fluctuations that are inherent with VAWTs. There is additional benefit in using the reaction torque provided by the generators to reduce the rotor speed at wind speeds above rated, so that the proportion of each revolution through which the flow over the blades is stalled can be increased, so limiting the VAWTs maximum power output.

As stated previously the only prior art that has been identified as relevant to the claims made for the present invention is the '1-180 Poseidon' Darrieus wind turbine proposed by Ljungstrom in 1980 (Proceedings of the Third International Symposium on Wind Energy Systems, Copenhagen, 1980, pp. 333-355), which also proposed the use of rollers to take the weight of the rotor and to deliver the rotor power output at a much increased speed to electrical generators. Ljungstrom's proposed wind turbine had a height of 210 m and a diameter of 180 m, with the curved 'egg-beater blades supported at the top and bottom by a central rotating tower which at ground (or sea) level had a diameter of 60 m. It was proposed that the rotor weight be supported by 56 rollers distributed around the periphery of the 60 m diameter base, 14 of which would be connected to electrical generators. The large downwind force on the rotor, acting at approximately mid-height, i.e. about 105 m above the base, would be reacted by a single bearing at ground level to take the downwind force but not the massive overturning moment. The only means of reacting this was to allow the downwind rollers to take most of the rotor weight whilst the upwind rollers were only lightly loaded. In fact the already very heavy rotor required added ballast to prevent the rotor from lifting off the upwind rollers. However a consequence of the very unequal rotor loading is that only the downwind rollers can be fully effective in transferring power to the generators, and most of the generators are operating inefficiently at low part load for most of the time. And like almost every other large VAWT built or proposed there is no practicable provision for stopping the rotor in an emergency.

Claims (19)

1. Claims 1. A large vertical axis wind turbine whose rotor comprises a horizontal disc-shaped hub to which is attached one or more substantially horizontal cross-arms each of which has a substantially vertical aerofoil section blade attached to its outer end, with the weight of this rotor supported by three or more radially-aligned and substantially horizontal cylindrical rollers (the drive rollers) whose diameter is small compared with the hub diameter, and each of these drive rollers is attached via its axle bearing housing to the non-rotating top of the wind turbine's tower and positioned directly underneath -and evenly spaced around a horizontal annular track (the drive track) which is centred on the rotor's vertical axis and fixed to the underside of the rotor hub so that as the rotor turns friction between this drive track and the drive rollers makes the latter rotate and deliver the wind turbine's power output by a shaft extension of each drive roller's axle to an electrical generator (either directly or via a gearbox) that is fixed to the tower top; the drive track diameter is similar to that of the hub and as the diameter of the drive rollers is much smaller they necessarily rotate at a speed that is much faster than the rotor speed, with the result that each drive roller's shaft delivers its power output to its associated electrical generator (either directly or via a gearbox) at a relatively high speed, with consequential substantial weight and cost benefits; an additional three or more cylindrical rollers (the reaction rollers) are also attached via their axle bearing housings to the wind turbine's tower but these have their axes vertical and are positioned in contact with and evenly spaced around the inside of a cylindrical track (the reaction track) which is also centred on the rotor's vertical axis and fixed to the rotor hub so that it rotates with it and these reaction rollers resist the large downwind force that necessarily results from the wind turbine's interaction with the wind.
2. 2. A wind turbine according to claim 1 in which two aerofoil-section blades are joined together at their root ends to make a double-length combination blade, and a combination blade as described is then attached to the outer end of each of the rotors one or more cross-arms such that the length of blade above the cross-arm is approximately the same as the length below; each of the one or more cross-arms has a length comparable with the length of the combination blade, and the rotor swept area is consequently much greater than that of a conventional horizontal axis wind turbine that uses blades of the same length as each of the two used to make a combination blade.
3. 3. A wind turbine according to any of the preceding claims in which at least part of the length of each cross-arm has a cross-section which is not only streamlined and low drag but also has an aerofoil profile such that as each cross-arm rotates around the rotor's vertical axis it experiences a substantial vertically upwards aerodynamic lift force, with the consequence that the airflow through the rotor experiences an equal but opposite downforce, pulling faster-moving air from above and ahead of the rotor into the rotor, thereby enhancing the rotor performance.
4. 4. A wind turbine according to any of the preceding claims in which a number of spring-applied and hydraulically-released brakes, with brake pads shaped to suit the curvature of the cylindrical reaction track, are fixed to the tower top structure and positioned around the inside circumference of the reaction track, interspersed between the reaction rollers, so as to provide means of braking the rotor when required.
5. 5. A wind turbine according to any of the preceding claims in which each drive roller is mounted via its axle bearing housing onto a rigid platform underneath the roller, and this platform is connected to the tower top structure using short-stroke hydraulic actuators such that each drive roller is able to move vertically through a short distance; several large-area flat and horizontal brake pads are also fixed to the tower top directly underneath the horizontal annular drive track, spaced around its circumference and interspersed between the drive rollers, so that when required or in an emergency the hydraulic system can be depressurised with the result that the rotor then drops under its own weight onto the brake pads and is brought to rest
6. 6. A wind turbine according to claim 5 in which the electrical generator and -if required -its associated gearbox is mounted on the same rigid platform as each drive roller (with this platform connected to the tower top structure using short-stroke hydraulic actuators) thereby avoiding the need for a flexible coupling between each drive roller and its gearbox (if required) and electrical generator.
7. 7. A wind turbine according to any of the preceding claims in which three or more additional radially aligned and substantially horizontal cylindrical rollers, the free rollers, are positioned directly beneath the drive track and interspersed between the drive rollers with each of these free rollers mounted via its axle bearing housing onto a rigid platform underneath it which is connected to the tower top structure using short stroke hydraulic actuators such that each free roller can be controlled to move vertically through a short distance, and this arrangement permits one or more of the drive rollers to be off-loaded by raising one or more of the adjacent free rollers until the drive roller is no longer in contact with the drive track (and if desired the drive roller can also be lowered); disconnecting one or more of the drive rollers in this way reduces transmission power losses when operating at part load and facilitates inspection and maintenance on individual drive rollers without necessitating rotor shut-down and without compromising the ability to automatically drop the rotor onto the substantially horizontal brake pads underneath the annular drive track when required or in an emergency by depressurising the hydraulic system.
8. 8. A wind turbine according to any of the preceding claims in which the drive rollers (as well as the free rollers, if utilised) are tapered, with their outer ends having a slightly larger diameter than their inner ends such that the circumferential speed at the surface of each roller matches the circumferential speed of the horizontal annular drive track at all positions along the length of each cylindrical roller.
9. 9. A wind turbine according to any of the preceding claims in which an additional means of preventing rotor overspeeding is provided by housing a drag parachute (such as is used by many light aircraft to provide a survivable descent in an emergency) within each cross-arm and located near its outer end, with its deployment triggered by a rotor overspeed sensor.
10. 10. A wind turbine according to claim 9 in which the drag parachute deployment is rocket assisted, with the parachute exiting the cross-arm through a frangible hatch in its lightweight tail section.
11. 11. A wind turbine according to any of the preceding claims in which the weight of each cross-arm and the combination blade attached to it is supported at least in part by one or more cables, each attached at their outer end to part of the cross-arm and with its inner end attached to or passing over a pylon fixed to the top of the rotor hub, thereby reducing bending stresses in the cross-arm and hence its weight.
12. 12. A wind turbine according to any of the preceding claims in which each cross-arm is made with a vertical pre-curve such that when the wind turbine is operating the cross-arm's outer end is approximately level with the mid-length position of the reaction rollers.
13. 13. A wind turbine according to any of the preceding claims in which resistive elements are housed within a suitable liquid-filled container fixed within the tower, such that if the electrical power output from the wind turbine is interrupted it can continue to operate without over-speeding by switching the power output to this resistive power dump; if the interruption is brief (up to a few seconds) the rotor can resume normal operation but if the interruption is sustained the resistive power dump can be used to allow electrical braking, supplemented by one or more of the braking systems provided.
14. 14. A wind turbine according to any of the preceding claims in which the horizontal annular drive track and the cylindrical reaction track are made from steel.
15. 15. A wind turbine according to any of the preceding claims in which the cylindrical drive rollers, the cylindrical free rollers (if utilised) and the cylindrical reaction rollers are made from steel.
16. 16. A wind turbine according to any of the preceding claims in which the hub main structure is a horizontal annular steel box beam, with a box cross-section that is rectangular, centred on the vertical axis of the rotor and with the flat and horizontal annular steel drive track fixed to its underside and with the cylindrical steel reaction track fixed to the inner cylindrical surface of the annular box beam.
17. 17. A wind turbine according to any of the preceding claims in which the flat annular drive track and the cylindrical reaction track are both made from a number of closely fitting, individually replaceable, segments.
18. 18. A wind turbine according to claim 17 in which the segments comprising the annular drive track are tapered in thickness with their inner ends thicker than their outer ends such that the angle that the drive track surface makes relative to the horizontal is equal to the angle between the axis of each drive roller and its rolling surface, so allowing the axis of each drive roller and the electrical generator it drives (either directly or through a gearbox) to remain horizontal.
19. 19. A wind turbine according to any of the preceding claims in which winglets are attached to the blade tips, so as to reduce the blade drag and enhance the wind turbine's efficiency.