Turbine with contrarotating rotors and associated method

Abstract

The disclosure relates to a turbine having contrarotating first and second rotors, and a method of operating such a turbine. The turbine comprises a generator having first and second generator parts. The first generator part comprises at least one electromagnet and is arranged to rotate with the first rotor. The second generator part comprises at least one electrical winding and is arranged to rotate with the second rotor. The method comprises controlling the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

Description

[0001] Turbine with contrarotating rotors and associated method

[0002] Field

[0003] The present disclosure relates to a turbine having at least two contrarotating rotors, and a method of operating such a turbine.

[0004] Background

[0005] The transition from non-renewable to renewable energy sources is an essential step in the fight against climate change. Examples of renewable energy sources include wind energy, hydropower and ocean energy. Wind energy is an especially popular renewable energy source: it is both abundant - it has the potential to meet a significant proportion of global electricity needs; and is present in almost all geographical regions of the world.

[0006] Wind turbines are well known devices for converting the kinetic energy of wind into electrical energy. Regardless of type, the general working principle is that turbines have a rotating part, called the rotor, connected to an electrical generator. When the wind blows, the rotor is caused to rotate. Rotation of the rotor causes rotation of a component of the electrical generator, which results in the generation of electricity.

[0007] Wind turbines may be considered to be horizontal or vertical axis. In a horizontal axis wind turbine (HAWT), the rotor rotates about a generally horizontally extending axis. Meanwhile, in a vertical axis wind turbine (VAWT), the rotor rotates about a generally vertically extending axis. In commercial wind farms, wind turbines are typically HAWTs. The HAWTs used in commercial farms typically have a single rotor with three blades arranged around the horizontal axis. When the wind blows, the blades (and thus rotor overall) are caused to rotate. Such HAWTs can be extremely large: some onshore commercial wind farm turbines have rotor diameters of over 150m, while some offshore commercial wind farm turbines have rotor diameters of over 200m. Meanwhile, VAWTs are typically smaller and can be usefully deployed in urban environments where the flow of the wind is less uniform (that is, more disturbed) than in the rural or offshore environments of commercial wind farms.

[0008] As the rate of climate change increases, the need to meet global electricity demands using renewable energy sources becomes ever more urgent. A widely used technique to increase electricity generation is to increase turbine rotor diameter size. In short, larger diameter rotors sweep a larger area, capture more kinetic energy, and so produce more electricity. However, with rotor diameters now extending beyond 150m, a ceiling is being approached which ultimately limits the upper rotor diameter. The ceiling exists for numerous reasons, including current legislative constraints and engineering constraints. Similar logic can be applied to the generation of electricity from other fluids, such as water.

[0009] An opportunity therefore exists to provide an alternative way of increasing the energy generated from renewable energy sources such as wind energy, hydropower and ocean energy. Summary

[0010] This overview introduces concepts that are described in more detail in the detailed description. It should not be used to identify essential features of the claimed subject matter, nor to limit the scope of the claimed subject matter.

[0011] According to a first aspect, there is provided a method of operating a turbine, the turbine comprising: a first rotor arranged to rotate about a rotor axis; a second rotor arranged to rotate about the rotor axis, the first and second rotors arranged for contrarotation about the rotor axis; a generator comprising first and second generator parts, wherein: the first generator part comprises at least one electromagnet and is arranged to rotate with the first rotor about the rotor axis; and the second generator part comprises at least one electrical winding and is arranged to rotate with the second rotor about the rotor axis; wherein the method comprises: determining a first parameter, wherein the first parameter is the rotational speed of one of the first rotor or the second rotor; determining a second parameter, wherein the second parameter is the rotational speed of the other of the first rotor or the second rotor; and, based on the determined first and second parameters, controlling the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

[0012] The method of the first aspect and the turbine of the second aspect have numerous advantages, which include the following. First, the provision of at least two, contrarotating rotors - rather than a single rotor - increases the energy captured and converted to electricity by a single turbine. This is at least because, for a given turbine size, the area of the working fluid (such as the air) swept by multiple rotors is larger than the area swept by a single rotor of the same size. As a result, more kinetic energy is captured and thus more electricity generated. Second, by integrating a single generator between the two rotors (rather than each rotor being connected to its own, dedicated generator), the total material cost - both financial and to the environment - is reduced. Third, despite both rotors being connected to a single generator, the turbine can still be controlled for maximum power output and safe operation by determining the rotational speeds of the first and second rotors and controlling the current in the at least one electromagnet and / or the at least one electrical winding accordingly. Fourth, by controlling the current to minimise the difference between the aerodynamic torque acting on each of the first and second rotors, maximum power point tracking strategies can be used for each rotor despite the two turbines not being connected to their own dedicated generators and whilst still maintaining safe operation of the turbine. Further, by actively minimising the difference between the aerodynamic torques, a condition in which both rotors rotate in the same direction (otherwise referred to in this disclosure as a stiction condition) is avoided.

[0013] The aerodynamic torque acting on each rotor may be a function of the rotational speed of the respective rotor and the fluid speed at the respective rotor.

[0014] Optionally, determining the second parameter comprises: determining the voltage in the at least one electrical winding; determining the first parameter; and determining the second parameter based on the determined voltage and the determined first parameter. Advantageously, this means that - at a minimum - only the voltage in the at least one electrical winding and information about the rotational speed of one of the rotors is needed to control the current in the electromagnet and / or electrical winding to deliver safe and efficient operation. Thus the apparatus may be technically simple which results in a reliable approach which reduces material usage.

[0015] Optionally, the method comprises determining one or more further parameters, wherein the one or more further parameters comprise one or more of: the direction of rotation of the first and / or the second rotor; the current in the at least one electrical winding; and / or the fluid (for example, air) speed at the turbine.

[0016] Optionally, the method comprises controlling the current in the at least one electromagnet and / or the at least one electrical winding based on one or more of the further parameters.

[0017] Optionally, the first parameter (that is, the rotational speed of the one of the first or second rotor) may be determined by directly measuring the rotational speed of the one of the first or second rotor, for example using a rotational sensor. Additionally or alternatively, the second parameter (that is, the rotational speed of the other of the first or second rotor) may be determined by directly measuring the rotational speed of the other of the first or second rotor, for example using the or a rotational sensor, or a further rotational sensor. In other words, the rotational speed of the first and / or second rotor may be determined directly (that is, directly measured) or indirectly.

[0018] Advantageously, determining the rotational speeds of the first and / or second rotor using rotational sensors improves accuracy. For example, this avoids any inaccuracies introduced due to hysteresis in the electromagnets when the rotational speeds are determined based on the measured rotational speed of one rotor and the voltage in the generator (that is, current in the electrical winding).

[0019] Optionally, the method comprises: detecting that the difference between the aerodynamic torque acting on each of the first and second rotors is above a (stiction) threshold; and, in response to detecting that the difference is above the (stiction) threshold, entering a stiction recovery operating mode.

[0020] Optionally, the method comprises, in the stiction recovery operating mode: decoupling the first and second generator parts by reducing the current in the at least one electromagnet to substantially zero. Advantageously, by reducing the current to substantially zero, there is no longer a magnetic interaction between the two generator parts and so the two rotors become unlocked.

[0021] Subsequently, or alternatively, the method may comprise, in the stiction recovery operating mode: operating the generator in a motoring mode. Advantageously, by operating the generator in the motoring mode, the direction of rotation of the generator is reversed which applies a braking effect to both rotors. Once both rotors are sufficiently slowed, the kinetic energy of the fluid (for example, air) is sufficient to restart the rotor rotating in the unintended direction in the correct direction, and thus for the stiction condition to be resolved. According to a second aspect, there is provided a turbine comprising: a first rotor arranged to rotate about a rotor axis; a second rotor arranged to rotate about the rotor axis, the first and second rotors arranged for contrarotation about the rotor axis; a generator comprising first and second generator parts, wherein: the first generator part comprises at least one electromagnet and is arranged to rotate with the first rotor about the rotor axis; and the second generator part comprises at least one electrical winding and is arranged to rotate with the second rotor about the rotor axis; measuring means for measuring a first parameter, wherein the first parameter is the rotational speed of one of the first rotor or the second rotor; the or further measuring means for measuring a second parameter, wherein the second parameter is: the rotational speed of the other of the first rotor or the second rotor; or the voltage in the at least one electrical winding; and a control system configured to: determine the rotational speed of the one of the first or second rotor based on the first parameter; determine the rotational speed of the other of the first or second rotor based on the second parameter; and control, based on the determined rotational speeds, the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

[0022] Optionally, when the second parameter is the voltage in the at least one electrical winding, the control system is configured to determine the rotational speed of the other of the first or second rotor based on the second parameter and the determined rotational speed of the one of the first or second rotor.

[0023] Optionally, the and the further measuring means are a single measuring means. That is, the same measuring means.

[0024] Optionally, the measuring means may directly or indirectly measure the first parameter. Additionally or alternatively, the or the further measuring means may directly or indirectly measure the second parameter.

[0025] Optionally, the turbine comprises additional measuring means configured to measure one or more further parameters, wherein the one or more further parameters comprise one or more of: the direction of rotation of the first and / or the second rotor; the current in the at least one electrical winding; and / or the fluid speed at the turbine.

[0026] Optionally, the control system is further configured to: determine one or more of the further parameters; and control the current in the at least one electromagnet and / or the at least one electrical winding based on the one or more further parameters.

[0027] Optionally, the measuring means comprises a rotational sensor configured to measure the rotational speed of the one of the first or second rotor. That is, to directly measure the rotational speed of the one of the first or second rotor.

[0028] Optionally, the or the further measuring means comprises a rotational sensor configured to measure the rotational speed of the other of the first or second rotor. That is, to directly measure the rotational speed of the other of the first or second rotor. Optionally, the control system is configured to monitor the difference between the aerodynamic torque acting on each of the first and second rotors. The control system may be further configured to, in response to detecting that the difference is above a (stiction) threshold, enter a stiction recovery operating mode.

[0029] Optionally, the control system is configured to, when in the stiction recovery operating mode: decouple the first and second generator parts by reducing the current in the at least one electromagnet to substantially zero. Subsequently, or alternatively, the control system may be configured, when in the stiction recovery operating mode, to operate the generator in a motoring mode.

[0030] Alternatively or additionally, the control system may be configured to: detect that the first and second rotors are rotating in the same direction about the rotor axis; and, in response to detecting that the first and second rotors are rotating in the same direction, enter a stiction recovery operating mode.

[0031] The first generator part may be carried by the first rotor. The second generator part may be carried by the second rotor. The turbine may comprise a third rotor arranged to rotate about the rotor axis.

[0032] The first generator part may be arranged to rotate with the first rotor in a geared manner, for example via a gearing system. Alternatively or additionally, the second generator part may be arranged to rotate with the second rotor in a geared manner, for example via a further gearing system. The gearing system and / or the further gearing system may be a mechanical or magnetic gearing system.

[0033] The generator may be an axial flux generator.

[0034] The rotational axis may extend in a direction parallel to the longitudinal direction of the turbine. For example, the turbine may be a VAWT. Alternatively, the rotational axis may extend in a direction perpendicular to the longitudinal direction of turbine. For example, the turbine may be a HAWT. Advantageously, when the turbine is a HAWT, because the turbine is a contrarotating turbine, rotational velocity imparted by one rotor on the fluid (for example, air) may be captured - or scavenged - by the other rotor. This increases the power extracted from the wind and so represents an efficiency improvement.

[0035] The turbine may be a wind turbine and the fluid may be air. Alternatively, the turbine may be a propeller and the working fluid may be a liquid, such as water.

[0036] The turbine may be arranged to perform the steps of the method of the first aspect.

[0037] The generator may be configured to operate in a or the motoring or generating regime.

[0038] The methods disclosed herein, including the method of the first aspect, may be performed by a or the control system, the control system comprising one or more computing devices. Accordingly, one or more computing devices configured to perform any of the methods disclosed herein is provided. Further, a computer program comprising instructions which, when the program is executed by one or more computing devices of the control system, cause the one or more computing devices to carry out any of the methods disclosed herein is provided.

[0039] Further, a computer-readable medium comprising instructions which, when executed by one or more computing devices of the control system(s), cause the one or more computing devices to carry out any of the methods disclosed herein is provided.

[0040] Optional features of one aspect may be optional features of another aspect (or example), as will occur to the skilled person.

[0041] Brief description of the figures

[0042] Illustrative implementations of the present disclosure will now be described, by way of example only, with reference to the drawings. In the drawings:

[0043] Figure 1 is a schematic cross-section view of part of a VAWT that is according to an aspect;

[0044] Figure 2 is a schematic cross-section view of part of a HAWT that is according to a further aspect;

[0045] Figure 3 is a flow chart showing steps of a method that is according to a still further aspect; and

[0046] Figure 4 is a block diagram showing a computing device suitable for carrying out the steps of the disclosed methods.

[0047] Detailed description

[0048] A turbine and method of operating such a turbine is provided which delivers an alternative way of increasing the energy generated from renewable energy sources. In more detail: the turbine has at least two rotors - a first and a second. The at least two rotors are arranged to rotate in different directions. The provision of two rotors increases the energy that can be captured and converted into electricity by the turbine. Additionally, the turbine has a single generator split across the two rotors. More specifically, a first part of the generator, which comprises an electromagnet, is arranged to rotate with one of the rotors. A second part of the generator, which comprises an electrical winding, is arranged to rotate with the other of the two rotors. Use of a single generator reduces the material cost (both financial and to the environment). Further still, the turbine has a control system configured to carry out method steps which include: (a) determining a first parameter, the first parameter being the rotational speed of one of the rotors; (b) determining a second parameter, the second parameter being the rotational speed of the other of the rotors; and (c) controlling the current in the electromagnet and / or electrical winding based on the determined first and second parameters and such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

[0049] The aerodynamic torque acting on a given rotor is a measure of the rotational force acting on that rotor as a result of the aerodynamic forces acting on the rotor. Importantly, the aerodynamic torque is a function of the rotor speed, and not only a function of the wind speed at the rotor.

[0050] Returning to the presently disclosed turbine and method, as a result of steps (a) to (c) recited above, maximum power point tracking strategies can be sought for each rotor despite the two turbines not being connected to their own dedicated generators and - crucially - whilst still maintaining safe operation of the turbine. Accordingly, overall turbine efficiency is improved. (For completeness, for every wind speed, there is an optimal blade rotation speed for maximum power extraction.) Further, by actively minimising the difference between the aerodynamic torques, a condition in which both rotors rotate in the same direction (otherwise referred to in this disclosure as a stiction condition) is naturally avoided. The turbine, method and technical advantages will now be discussed in more detail in reference to the figures.

[0051] An example of an aspect of the present invention which addresses the problems of existing approaches is shown in Figure 1. Figure 1 shows part of a VAWT 100. More specifically, Figure 1 shows a simplified schematic cross-section view of a VAWT 100. The structure, operation and advantages of the VAWT 100 will now be described in relation to Figure 1.

[0052] The VAWT 100 has a vertically extending shaft 105 which supports a first rotor 110 and a second rotor 120. The first 110 and second 120 rotors are arranged for rotation about a rotor axis 101. In this example, the rotor axis 101 is common to both rotors 110, 120; however in other examples, this may be different. The rotor axis 101 is in this example parallel to, and coincident with, the vertically extending shaft 105. The first 110 and second 120 rotors are arranged for contrarotation about the rotor axis 101.

[0053] In this example, the first rotor 110 is the uppermost rotor. The first rotor 110 comprises a first rotor central hub 111 and, in this example, three blades 112 (hereafter referred to as first rotor blades 112). Two of the first rotor blades 112 are shown in Figure 1 . Each of the plurality of first rotor blades 112 is connected to the first rotor central hub 111 via a respective connecting member 113 (hereafter referred to as first rotor connecting members 113). In this example, each first rotor blade 112 comprises a generally vertically extending aerofoil section. Each first rotor connecting member 113 extends between its respective first rotor blade 112 and the first rotor central hub 111 such that the respective first rotor blade 112 is offset away from the central hub 111 in both horizontal and vertical directions, and such that the respective first rotor blade 112 extends upwards in the vertical direction. That is, the first rotor blades 112 extend up and away from the first rotor central hub 111. The first rotor central hub 111 is attached to the vertically extending shaft 105 (as described in more detail below).

[0054] The first rotor central hub 111 carries on its underside an upper mount 114. The upper mount 114 is arranged for rotation with the first rotor central hub 111 (and thus the first rotor 110 overall). The upper mount 114 is in this example disc shaped; however, in alternative arrangements, the upper mount may be differently shaped. In this example, the axis of rotation of the disc-shaped upper mount 114 is coincident with the rotor axis 101 , however other arrangements may be adopted. The upper mount 114 can be considered as a passive component. The upper mount 114 carries on its underside a first generator part 131. Rotation of the first rotor 110 thus also causes rotation of the first generator part 131. The first generator part 131 is attached to the upper mount 114 by, for example, bolts. Other attachment approaches may alternatively or additionally be used, such as welding and / or using adhesive. The first generator part 131 is an electromagnet. As the skilled person would appreciate, in an electromagnet, wire is wound into coils. When a current flows through the wire, a magnetic field is created. The strength of the magnetic field created is proportional to the strength of the current flowing through the wire. Although not explicitly shown, the first generator part 131 includes slip rings. Although not shown in detail, the first generator part 131 - and specifically, the wire(s) in the electromagnetic - are connected to an electricity supply via the slip rings such that, when desired, a current may be caused to flow in the wire(s) and a magnetic field be established.

[0055] Turning now to the second rotor 120, the second rotor 120 is substantially the same as the first rotor 110, but is inverted about a substantially horizontal axis. That is, the second rotor 120 is substantially a mirror image of the first rotor 110. The second rotor 120 is arranged beneath the first rotor 110. In this way the second rotor 120 is the lowermost rotor. Like the first rotor 110, the second rotor 120 comprises a second rotor central hub 121 and, in this example, three blades 122 (hereafter referred to as second rotor blades 122). Two of the second rotor blades 122 are shown in Figure 1. Each of the plurality of second rotor blades

[0056] 122 is connected to the second rotor central hub 121 via a respective connecting member

[0057] 123 (hereafter referred to as second rotor connecting members 123). In this example, each second rotor blade 122 comprises a generally vertically extending aerofoil section. Each connecting member 123 extends between its respective second rotor blade 122 and the second rotor central hub 121 such that the respective blade 122 is offset away from the central hub 121 in both horizontal and vertical directions, and such that the respective second rotor blade 122 extends downwards in the vertical direction (and thus down and away from the second rotor central hub 121). The second rotor central hub 121 is attached to the vertically extending shaft 105 (as described in more detail below).

[0058] The second rotor central hub 121 carries on its topside a lower mount 124. The lower mount

[0059] 124 is arranged for rotation with the second rotor central hub 121 (and thus the second rotor 120 overall). The lower mount 124 is substantially the same, and arranged in the same way, as the upper mount 114 and thus may likewise be disc or otherwise shaped.

[0060] The lower mount 124 may be considered to be a passive component. The lower mount 124 carries on its topside a second generator part 135. Rotation of the second rotor 120 thus also causes rotation of the second generator part 135. The second generator part 135 is attached to the lower mount 124 by, for example, bolts. Other attachment approaches may alternatively or additionally be used, such as welding and / or using adhesive. The second generator part 135 is an electrical winding. Although not explicitly shown, the second generator part 135 includes slip rings. As the skilled person would well appreciate, the second generator part 135 - and specifically, the electrical winding - is connected to an electricity supply via the slip rings such that a current may flow into or from the electrical winding. The first 131 and second 135 generator parts together form the VAWT generator 130, the generator 130 being arranged therefore between the first 110 and second 120 rotors. Though not shown in Figure 1 , the vertically extending shaft 105 extends through - and is attached to - the centres of the second rotor central hub 121 , the second rotor lower mount 124, and the first rotor upper mount 114. In this example, the first rotor central hub 111 is attached to the top of the vertically extending shaft 105 however other physical arrangements are possible. For example, the vertically extending shaft 105 may extend partially or wholly through the centre of the first rotor central hub 111.

[0061] Also not shown, the vertically extending shaft 105 has an upper section and a lower section. The upper and lower sections are arranged to rotate about the rotor axis 101. Specifically: the upper section is attached to, and is arranged to rotate with, the first rotor 110; and the lower section is attached to, and is arranged to rotate with, the second rotor 120. The upper and lower sections are arranged to rotate independently of each other. As a result, the upper and lower sections may rotate in opposite directions. This is achieved via linkage of the upper and lower sections via a centre bearing 140. The centre bearing 140 is arranged to maintain the upper and lower sections in a coaxial arrangement, whilst still allowing for contrarotation. In this example, the centre bearing 140 is a tapered roller bearing, however other arrangements - for example, other bearing types - are possible.

[0062] Also not shown in Figure 1 , the turbine 100 includes means for measuring the rotational speed of the first 110 and second 120 rotors (that is, measuring means). In this example, the measuring means comprises a first DC motor arranged on the upper section of the shaft 105 and a second DC motor arranged on the lower section of the shaft 105. The rotational speed of the upper section of the shaft 105, and thus of the first rotor 110, can be inferred from the voltage induced in the first DC motor by rotation of the upper section (and thus first rotor 110). Meanwhile, the rotational speed of the lower section of the shaft 105, and thus of the second rotor 120, can be inferred from the voltage induced in the second DC motor by rotation of the lower section (and thus second rotor 110). In this way, in the described example, the rotational speeds of the first 110 and second 120 can be considered to be indirectly measured.

[0063] In alternative arrangements, the rotational speed of one or the first 110 or second 120 rotors may be indirectly measured in the following way. The turbine 100 may include either the first or the second DC motor (or an alternative). As such, the rotational speed of the corresponding one of either the first 110 or second 120 rotor may be indirectly measured. Meanwhile, the rotational speed of the other may be determined from the voltage induced in the electrical winding together with the measured rotational speed of the other rotor. This is described in more detail below. The turbine 100 may thus, additionally or alternatively to the first and / or second DC motor(s), include means for measuring the voltage induced in the electrical winding.

[0064] Additionally or alternatively, the measuring means may comprise one or more rotational sensors, each rotational sensor configured to directly measure the rotational speed of one of the rotors. Using rotational sensors improves accuracy by, for example, avoiding inaccuracies introduced due to hysteresis. As will occur to the skilled person, one or both of the DC motors described above may be replaced by a rotational sensor. Other means for measuring - directly or indirectly - the first and second parameters may also be used, as will occur to the skilled person. Notably, in line with the above example, one of the first and second parameters may be measured (for example, directly via a rotational sensor, or indirectly via DC motor), while the other may be determined from the measured one of the first or second parameters in combination with the voltage induced in the electrical winding.

[0065] Also not shown in Figure 1 , the turbine 100 includes a control system which may comprise a computing device. An example computing device is described in relation to Figure 4. As discussed in more detail in reference to Figure 4, the computing device may, for example, include a memory and processor, with instructions stored in the memory which cause the processor to carry out the method steps described in relation to Figure 3 (Figure 3 being discussed in more detail later in this disclosure). The computing device may be arranged in a unit mounted on the shaft 105, in a unit mounted near the shaft 105, or in a unit remote from the turbine 100. The components of the computing device may additionally be distributed with, for example, some components arranged near or on the turbine 100, and some arranged remote from the turbine 100. Various arrangements are thus possible.

[0066] In alternative arrangements, there may be a different number of rotor blades 111 , 121. For example, there may be fewer blades (for example, two); or more blades (for example, five) blades. The first 110 and second 120 rotors may have a different number of blades.

[0067] In alternative arrangements, the blades 111 , 121 may have a different shape to that described. For example, the blades may be aerofoil sections which have a substantially spiral shape, the spiral having a vertically extending component such that each blade has approximately the shape of a single helix.

[0068] The first generator part 131 may comprise a plurality of electromagnets. The second generator part 135 may comprise a plurality of electrical windings.

[0069] Optionally, the turbine 100 may include one or more means for measuring wind speed at each rotor. The means may be, for example, one or more anemometers. The wind speed measurements may be used to control the turbine 100 and may increase the accuracy, easy and speed of control.

[0070] Further optionally, one or both of the generator parts may be arranged to rotate with the respective rotor in a geared manner. This may be, for example, via a mechanical and / or a magnetic gearing system. This is in contrast to a direct drive arrangement.

[0071] Example structural arrangements for the first 131 and / or second 135 generator parts will now be described. The first 131 and second 135 generator parts may have approximately the same structural arrangement, or have different structural arrangements. The example structural arrangements include:

[0072] 1 . Coreless electromagnetic arrangement. In this arrangement, no (or substantially no) ferromagnetic material is provided in the generator part. This approach has the advantage of being simple to manufacture and keeps use of materials to a minimum, thus having good environmental credentials. 2. Electromagnetic arrangement with ferromagnetic cores. In this arrangement, each winding of the electromagnet is arranged around a ferromagnetic core. The ferromagnetic cores may be, for example, laminations of insulated steel. This approach has the advantage that the laminations of insulated steel: (a) enhance the magnetic field created by the current flowing in the windings due to steel’s ferromagnetic properties; and (b) the insulation between the laminations has the effect of reducing eddy current losses and so improving efficiency. Alternatively or additionally, the ferromagnetic cores may be, for example, solid steel cores.

[0073] 3. Electromagnetic arrangement with magnetic-particle epoxy composite. In this arrangement, the windings of the electromagnet are embedded in an epoxy resin containing magnetic particles. The magnetic particles may be iron particles. For example, the epoxy resin may be produced by mixing magnetic particles in a thermosetting or thermoplastic polymer. Advantageously, this approach means that: (a) the magnetic field created by the current flowing in the windings is enhanced due to the magnetic particles’ ferromagnetic properties; (b) the epoxy composite has the effect of reducing eddy current losses and so improving efficiency; (c) the characteristics of the generator part can be tuned by varying the amount of the magnetic powder in the epoxy; and (d) by using an epoxy resin, the geometry of the magnetic-particle epoxy composite component(s) can be easily varied, and non- orthogonal shaped components formed. For example, wedge-shaped components can easily be formed. This is because the polymer is initially a liquid, so can flow into an arbitrary geometry before being set into a solid form.

[0074] In an example practical solution for an iron-particle epoxy composite, 200 ml of resin may be mixed with 200 ml of hardener (in this case, the hardener being the catalyst for the polymerisation process in the plastic formation). Immediately after adding the hardener, 1kg of iron powder with a purity of more than 99% iron and powder particle size of 45 pm for around 80% of the particles may be added to the resin hardener mix and mixed in. The resin may then be poured into a mould of the desired shape and allowed to harden. Other practical solutions for forming a metallic-particle epoxy composite may be used too.

[0075] Overall, the suitability of the above described arrangements to the first 131 and second 135 generator parts advantageously allows for such selections to be made that tune the key generator characteristics to each specific turbine application. Such generator characteristics include the cogging torque, flux density, winding inductance, and iron losses.

[0076] Operation of the VAWT 100 will now be described. In use, a current is applied to the electromagnet comprised in the first generator part 131 , as described above. When the turbine 100 experiences sufficiently strong wind, the first 110 and second 120 rotors are caused to rotate about the rotor axis 101. In normal operating conditions, the first 110 and second 120 rotors are caused to rotate in opposite directions due to the orientation of their blades 112, 122. That is, the first 110 and second 120 rotors are caused to contrarotate. Rotation of the first rotor 110 results in rotation of the first generator part 131 carrying the electromagnet. Rotation of the electromagnet results in a rotating magnetic field relative to the electrical winding of the second generator part 135. As will be well understood by the skilled person, the intersection of the electrical winding with the rotating magnetic field (and corresponding flux) generates an electromotive force (EMF) in the electrical winding. This causes a current to flow in the electrical winding. Power is thus generated by the turbine 100, and the generated power may be fed out from the turbine 100 in the usual way and may be fed to, for example, a battery, a main electricity supply grid or an industrial process. For example, the generated power may be conditioned using power electronics such that the power is fed out at a desired voltage and frequency. For example, at a voltage and frequency suitable for integration into the main electricity supply grid.

[0077] Advantageously, it is a distinct benefit of the present invention that the power output from the VAWT 100 is maximised whilst steady state operation is maintained, and despite there being only a single generator 130. This will now be described in more detail.

[0078] In a contrarotating rotor, the maximum power output for one rotor may be achieved at a different rotational speed than the other rotor .This is due to, for examples, changes in wind conditions. In a contrarotating turbine in which each rotor is attached to its own, dedicated generator, the speed of each rotor may be independently controlled. In this way, the maximum power output for each rotor may be achieved. For completeness, the reason the speed of each rotor may be independently controlled is because the effective load on each generator can be varied independently. As a result, the torque exerted by each generator against the motion of its respective rotor can be controlled independently. This allows for the speed of the respective rotor to be independently controlled. Corrections can also be made to account for the aerodynamic coupling between the rotors.

[0079] However, in the turbine 100, a single generator 130 is used and shared between the first 110 and second 120 rotors. As a result, such independent control of each rotor’s speed is no longer possible. In the present invention, this problem is solved by - rather than seeking to control each rotor’s speed - the generator 130 is operated such that the difference between the aerodynamic torque acting on each of the first 110 and second 120 rotors is minimised, whilst also seeking maximum power out. As discussed above, the aerodynamic torque is the torque generated by the force of the wind acting on the blades of the turbine. It is thus the aerodynamic torque which causes the turbine blades to rotate and, in turn, which causes rotation of the generator parts. The aerodynamic torque originates from the aerodynamic interaction between the wind and the blades, which enables the conversion of wind energy to mechanical (kinetic) energy. The aerodynamic torque acting on the blades is not only a function of the fluid properties of the wind at the blades (e.g., the wind speed), but also a function of factors such as the rotor speed and the blade pitch angle. Advantageously, minimising the difference between the aerodynamic torque acting on the first 110 and second 120 rotors maintains the turbine 100 in a steady (stable) state of operation, whilst also allowing for maximum power output to be sought.

[0080] In more detail, minimisation of the difference between the aerodynamic torque values is achieved by: (a) determining the rotational speeds of the first 110 and second 120 rotors; (b) determining from those speeds the aerodynamic torque exerted on each of the first 110 and second 120 rotors; and (c) and controlling the current in the electromagnet (or electromagnets if more than one is present) carried by the first generator part 131 and / or the current in the electrical winding (or electrical windings, if more than one is present) carried by the second generator part 135 such that the difference between the aerodynamic torque acting on each of the first 110 and second 120 rotors is minimised. Accordingly, an optimisation strategy is carried out to optimise (specifically, minimise) the difference between the aerodynamic torques exerted on the first 110 and second 120 rotors. An example optimisation strategy is detailed later in this disclosure.

[0081] Advantageously, maximum power output can still be sought by, for example, increasing the current in the electromagnet(s). However, if this leads (or would lead) to an increase, or an increase above a threshold amount, in the difference between the aerodynamic torques on the first 110 and second 120 rotors, then this change is reversed (or rejected, if not yet made) or further adjustments in the system made to decrease the difference between the aerodynamic torques. In other words, it is an aim of the present control strategy that the difference between the aerodynamic torques on the first 110 and second 120 rotors remains below a threshold amount. It may also be an aim of the present control strategy that the difference remains approximately constant and at a level below the threshold amount. This may otherwise be referred to as a numerical target (for the difference between the aerodynamic torques). The numerical target may additionally or alternatively be used with or to the threshold amount. Advantageously, having a numerical target and / or threshold amount helps avoid a stiction condition developing. This is because increasing the difference in aerodynamic torques at any point makes entering stiction in the future more likely. Thus any increase is undesirable, even if such an increase would not result in a stiction condition.

[0082] Optionally, operation of the turbine 100 may also include controlling the generator torque. The generator torque is the torque exerted on the rotating components (i.e., the first 110 and second 120 rotors) due to currents flowing in the electrical winding(s) of the generator (i.e., the electrical winding(s) carried by the second generator part 135). The generator torque acts against the motion of the generator. That is, it acts to decelerate the relative motion of the contrarotating parts. Controlling of the generator torque is discussed in more detail below.

[0083] As mentioned above, in a conventional, single rotor turbine, the wind speed can be assumed to be approximately constant. From aerodynamic considerations, the power extracted from the wind can thus be taken as varying with the rotational speed of the rotor. For a given single rotor turbine, there is a particular rotational speed when the power extracted is at its maximum. Thus, in a single rotor turbine, it is desirable to control the rotational speed of the rotor such that it allows for maximum power extraction. This can be achieved by varying the generator torque, which in turn varies the rotational speed of the rotor. For example, if the generator torque is larger than the aerodynamic torque, then the rotor speed decreases, and vice versa. In practice, the generator torque is commonly set to a value which results in a steady state at the optimum rotor speed. This concept broadly holds true for a turbine having more than one rotor, such as a dual rotor turbine. If separate generators were provided for each rotor, then - given there is an optimum rotational speed for each rotor of the turbine to give maximum power out - each generator would be controlled to provide the necessary generator torque to achieve the optimum rotational speeds. However, in the turbine 100, this is not possible due to the presence of a shared generator 130. Due to the use of a shared generator 130 in the turbine 100, the same generator torque is exerted by the generator 130 against the motion of both rotors 110, 120. Thus in the present example, the generator 130 is controlled to provide optimum relative rotational speed between the two rotors (in order to seek maximum power out); but is controlled not to the detriment of balancing the aerodynamic torque experienced by each rotor.

[0084] Notably, a distinct advantage of the contrarotating arrangement of the turbine 100 is that, because the first 131 and second 135 generator parts are both spinning, the relative rotational speed between them is increased compared to in a conventional turbine (when one part would be stationary). This results in higher power output capabilities.

[0085] Further advantageously, by controlling the turbine 100 as set out above, a stiction condition is naturally avoided. This is because, as explained, controlling the turbine 100 to minimise the aerodynamic torque difference inherently promotes operation of the turbine 100 in the steady state, and thus avoids a stiction condition. However, if the difference between the aerodynamic torques acting on each of the first 100 and second 120 rotors in any case increases, the turbine 100 is controlled such that, if the difference is detected to be above a (stiction) threshold, in response, a stiction recovery operating mode is entered. In the stiction recovery operating mode, the first 131 and second 135 generator parts are decoupled. This can be achieved by reducing the current in the electromagnet(s) to substantially zero. As a result, there is no longer a magnetic interaction between the first 131 and second 135 generator parts and so the two rotors become unlocked. That is, decoupled. To return to steady state operation, optionally the generator 130 may be operated in a motoring mode. For example, by firing reset pulses on the side of the second generator part 135. Usefully, this reverses the direction of the generator 130 which applies a braking effect to both the first 110 and second 120 rotors. Once both rotors 110, 120 are sufficiently slowed, the kinetic energy of the fluid - for example, air - is sufficient to restart the rotor 110, 120 that was rotating in the unintended direction in the correct direction. As a result, the stiction condition is advantageously resolved and a steady state of operation returned to.

[0086] To conclude, it is an inherent advantage of the turbine 100 and control strategy for the turbine 100 that stiction is avoided (by active aerodynamic torque management). If stiction occurs however, the turbine 100 can be operated to: stop the condition (by decoupling the generator parts), and optionally actively recover the turbine 100 by operating the generator 130 in a motoring mode.

[0087] Optionally, the turbine 100 may additionally include means for measuring the direction of rotation of the first 110 and / or second 120 rotors. This direction information can additionally or alternatively be used to determine a stiction condition.

[0088] A stiction condition may alternatively or additionally be detected by determining: (a) that both rotors 110, 120 are moving (from non-zero rotational speeds, for example); and (b) that, despite this, power output from the generator 130 is substantially zero. This indicates that the rotors 110, 120 are rotating in the same direction.

[0089] Additionally or alternatively, in the stiction recovery operating mode, the rotors 110, 120 may be braked by one or more of the following techniques:

[0090] • changing the pitch of the first rotor blades 112 and / or second rotor blades 122; • having a mechanical brake such as callipers gripping a disc brake on the shaft of the first 110 and / or second 120 rotor; and / or

[0091] • short circuiting the generator load, for example by applying a maximum current to the electromagnet(s).

[0092] An example of a further aspect of the present invention, which likewise addresses the problems of existing approaches, is shown in Figure 2. Figure 2 shows part of a HAWT 200. More specifically, Figure 2 shows a simplified schematic cross-section view of a HAWT 200. The structure, operation and advantages of the HAWT 200 will now be described in relation to Figure 2.

[0093] There are corresponding features between the VAWT 100 of Figure 1 and the HAWT 200 of Figure 2. Corresponding parts are presented with like reference numbers, but increased by 100. For example, the first rotor in the VAWT 100 of Figure 1 has reference number 110, meanwhile the corresponding first rotor in the HAWT 200 of Figure 2 has reference number 210. For brevity, corresponding parts in the HAWT 200 are not described in detail. Instead, an overview is provided below which focuses on the differences between the VAWT 100 and HAWT 200.

[0094] Like the VAWT 100, the HAWT 200 has a vertically extending axis 205 which supports a first rotor 210, a second rotor 220, and a generator 230. The first 210 and second 220 rotors are arranged for rotation about a rotor axis 201. The rotor axis 201 is in this example perpendicular to the vertically extending shaft 205. That is, the rotor axis 201 is substantially horizontal. The first 210 and second 220 rotors are arranged for contrarotation about the rotor axis 201 on respective shafts (first rotor shaft 213 and second rotor shaft 223). In this example, the first 213 and second 223 rotor shafts are arranged generally above, and approximately equidistant from, the vertically extending shaft 205. In other arrangements, the physical arrangement in space of the first 213 and second 223 rotor shafts (and indeed rotors 210, 220) may be different.

[0095] In this example, the first rotor 210 is shown in Figure 2 as being arranged on one side (specifically, the left hand side) of the vertically extending shaft 205 and the second rotor 220 on the other side (specifically, the right hand side) of the vertically extending shaft 205. This is by way of example only however and the arrangement of the rotors 210, 220 either side of the vertically extending shaft 205 may be flipped.

[0096] The first rotor 210 comprises a first rotor central hub 211 and, in this example, three blades

[0097] 212 (hereafter referred to as first rotor blades 212). Two of the first rotor blades 212 are shown in Figure 2. Each of the plurality of first rotor blades 212 is connected to, and equally spaced around, the first rotor central hub 211. In this example, each first rotor blade 212 comprises an aerofoil section which is longer in the chord direction than it is thick. The first rotor central hub 211 is attached to a first end of the first rotor shaft 213. The first rotor shaft

[0098] 213 is substantially horizontally arranged, and thus is generally perpendicular to the vertically extending shaft 205. The rotor axis 201 is coincident with the first rotor shaft 213. The first rotor shaft 213 comprises the first end (attached to the first rotor central hub 211) and a second end. The second end is arranged closer to the vertically extending shaft 205 than the first end. Attached to the second end of the first rotor shaft 213 is a first rotor mount 214. The first rotor mount 214 is arranged for rotation with the first rotor shaft 213, the first rotor central hub 211 , and thus the first rotor 210 overall. The first rotor mount 214 is in this example generally disc shaped, the axis of rotation of the disc being arranged generally parallel and coincident with the rotor axis 201. However, in alternative arrangements, the first rotor mount 214 may be differently shaped or arranged. As in the VAWT 100, the first rotor mount 214 can be considered as a passive component. The first rotor mount 214 carries on its side which is closer to the vertically extending shaft 205 (i.e. , its vertically extending shaft 205 facing side), a first generator part 231. Rotation of the first rotor 210 thus causes rotation of the first generator part 231. The first generator part 231 is generally as described in relation to the first generator part 131 of the VAWT 100 of Figure 1.

[0099] Turning now to the second rotor 220, the second rotor 220 is substantially the same as the first rotor 210, but is inverted about the vertically extending shaft 205. That is, the second rotor 220 is substantially a mirror image of the first rotor 210. Like the first rotor 210, the second rotor 220 comprises a second rotor central hub 221 and, in this example, three blades 222 (hereafter referred to as second rotor blades 222). Two of the second rotor blades 222 are shown in Figure 2. The second rotor blades 222 are substantially the same as the first rotor blades 212. The second rotor central hub 221 is attached to a first end of the second rotor shaft 223. The second rotor shaft 223 is substantially horizontally arranged, and thus is generally perpendicular to the vertically extending shaft 205. The rotor axis 201 is coincident with the second rotor shaft 223. The second rotor shaft 223 comprises the first end (attached to the second rotor central hub 221) and a second end. The second end is arranged closer to the vertically extending shaft 205 than the first end.

[0100] Attached to the second end of the second rotor shaft 223 is a second rotor mount 224. The second rotor mount 224 is arranged for rotation with the second rotor shaft 223, the second rotor central hub 221 , and thus the second rotor 220 overall. The second rotor mount 224 is in this example generally the same as the first rotor mount 214. However, in alternative arrangements, the second rotor mount 224 may be different to the first rotor mount 214 and / or have a different arrangement to that shown and described in relation to Figure 2.

[0101] As in the VAWT 100, the second rotor mount 224 may be considered to be a passive component. The second rotor mount 224 carries on its side which is closer to the vertically extending shaft 205 (i.e., its vertically extending shaft facing side), a second generator part 235. Rotation of the second rotor 220 thus also causes rotation of the second generator part 235. The second generator part 235 is attached to the second rotor mount 224 in the same way, or in the alternative ways, described in relation to the second rotor lower mount 124 described in relation to Figure 1. The second generator part 235 comprises an electrical winding. The first 231 and second 235 generator parts together form the HAWT generator 230, the generator 230 being arranged therefore between the first 210 and second 220 rotors.

[0102] In the example shown in Figure 2, the generator 230 is arranged above, and approximately centrally arranged upon, the vertically extending shaft 205; however other physical arrangements are possible. For example, the generator 230 need not be arranged directly above the vertically extending shaft 205, but instead may be displaced horizontally from the vertically extending shaft 205. A notable difference between the VAWT 100 and the HAWT 200 arrangements is that, in the HAWT 200, no centre bearing 140 is needed between the first and second rotors. Instead, in this example, there is a small air gap between the first 210 and second 220 rotors. For completeness, the bearing is not required in the HAWT 200 because the first 213 and second 223 rotor shafts are separate components supported in space (directly or indirectly by the vertically extending shaft) and the two shafts do not necessarily need to interact; whereas in the VAWT 100, a single vertically extending shaft 105 extends through the turbine 100, with two sections (the upper and lower previously described in relation to Figure 1) arranged for contrarotation. Hence the centre bearing 140 is required in the VAWT 100 but is not essential in the HAWT 200. In alternative arrangements, the HAWT 200 may however still comprise a centre bearing (otherwise referred to as an alignment bearing), such as a bush with integrated needle roller bearing. Advantageously, the provision of a centre bearing in the HAWT 200 arrangement can help with the alignment of the first 213 and second 223 rotor shafts.

[0103] As for the generator 130 of the VAWT, although not shown in Figure 2, the generator 230 of the HAWT 230 includes slip rings arranged such that current be supplied to (or drawn from) the electromagnet and electrical winding.

[0104] Although not shown in Figure 2, the turbine 100 includes means for measuring the rotational speed of the first 210 and second 220 rotors as described in relation to Figure 1. As a result, in the way described in relation to Figure 1 , the rotational speeds of the first 210 and second 220 can be considered to be directly measurable.

[0105] Also not shown in Figure 2, the turbine 200 includes a control system as described in relation to Figure 1.

[0106] Although the first rotor blades 212 and the second rotor blades 222 are shown in Figure 2 as being substantially the same, in alternative arrangements, the first 212 and second 222 rotor blades may be different. In particular, the first rotor blades 212 may be differently sized and / or shaped to the second rotor blades 222. For example: one of the rotors may be considered to be the downwind rotor and that rotor configured to have blades of larger diameter than the blades of the other rotor. The other rotor may be considered to be the upwind rotor. Advantageously, the different sizing of the blades accounts for the fact that the downwind rotor will receive less input power from the fluid than the upwind rotor. The rotor blades of the downwind rotor may be for example approximately double the diameter of the rotor blades of the upwind rotor. Advantageously this means that - in the steady state of operation - the torques on the two rotors are approximately equal. This makes the control strategy described in this disclosure even more effective.

[0107] As the skilled person will understand, many of the alternative arrangements described in relation to the VAWT 100 are equally applicable to the HAWT 200. For brevity, those alternatives will not be repeated here but should be considered to apply where technically compatible.

[0108] Operation of the HAWT 200 is substantially the same as described in relation to the VAWT 100. In addition however to the advantages described in relation to operation of the VAWT 100, there is a further distinct operational benefit of the HAWT 200. This will now be described. In a single rotor HAWT, a component of rotational velocity is imparted to the fluid leaving the rotor blades. This is sometimes referred to as swirl. In a single rotor HAWT, this rotational motion represents a source of wasted kinetic energy that is carried away from the turbine unharvested. In the HAWT 200 however, the presence of (at least) a second rotor means that this usually wasted kinetic energy is harvested and converted to kinetic energy by the second rotor, and subsequently converted to electrical energy by the generator 230. Thus the multi rotor HAWT 200 of the present invention offers improved energy harvesting performance over single rotor HAWTs.

[0109] Figure 3 shows steps of a method 300 that is according to a still further aspect. As discussed, the VAWT 100 and the HAWT 200 previously described may be arranged to perform the steps of the method 300. In particular, the turbines 100, 200 may include the previously described control system having a memory in which instructions are stored which, when executed by a processor of the control system, result in the steps of the method 300 being executed. Each of the steps of the method 300 will now be described.

[0110] The method 300 initiates in this example at step 310. At step 310, the step of determining a first parameter is carried out, wherein the first parameter is the rotational speed of the first rotor or the second rotor. As discussed in relation to the VAWT 100 and the HAWT 200, the rotational speed may be determined directly (for example, via measuring using a rotational sensor); or indirectly (for example, via voltage induced in a DC motor attached to the shaft of the rotor being investigated).

[0111] Next, at step 320, the step of determining a second parameter is carried out, wherein the second parameter is the rotational speed of the other of the first rotor or the second rotor. Once more, the rotational speed may be determined directly or indirectly. In particular, the second parameter may be determined indirectly from the voltage induced in the at least one electrical winding in combination with the rotational speed of one of the rotors. From this, the rotational speed of the other of the rotors can be determined. Advantageously, this reduces the measurement apparatus required, resulting in a simple setup with reduced material usage and good reliability.

[0112] Next, at step 330, the step of controlling, based on the determined first and second parameters, the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised is carried out. It follows that step 330 may include determining the aerodynamic torque acting on each of the first and second rotor.

[0113] The advantages of controlling the turbine 100, 200 to balance the aerodynamic torques acting on each of the rotors is as previously described.

[0114] Once step 330 is complete, the method may return to step 310. Further still, multiple iterations of the method 300 may be running in parallel. For example, while step 330 is being carried out for a first iteration - that is, for particular rotational speed values, step 310 may be being carried out for a subsequent iteration. Optionally, the method 300 may also include the following stiction detection and recovery steps. At step 340, the method 300 may include detecting that the difference between the aerodynamic torque acting on each of the first and second rotors is above a (stiction) threshold. The (stiction) threshold may be as described below. In response to detecting that the difference is above the (stiction) threshold, a stiction recovery operating mode may be entered into.

[0115] In one example, the stiction threshold may be a discrete value. In another example, the stiction threshold may be percentage based: for example, the stiction threshold may be based on the aerodynamic torque acting on one rotor being a percentage, or below a percentage, of the aerodynamic torque acting on the other rotor. The percentage may be for example 25%, and thus the threshold may be met if the aerodynamic torque acting on either rotor is 25% or less than the aerodynamic torque acting on the other rotor. In any case, the precise approach and value taken for the stiction threshold will vary dependent upon the location of the wind turbine. For example, the threshold may be different for a turbine in an urban location with more turbulent fluid (where the stiction condition should be more conservative), versus a turbine in a rural location with less turbulent winds.

[0116] In another example, the stiction threshold may be met if the aerodynamic torque acting on either rotor is zero, or has the opposite direction to that expected in the steady state.

[0117] For completeness, alternatively or additionally, a simple method for stiction condition detection would be to detect if the rotational speed of either rotor is zero (or if either rotor is rotating in the opposite direction to that expected in the steady state).

[0118] In the stiction recovery operating mode, the method 300 may include step 350, in which: the first and second generator parts are decoupled by reducing the current in the at least one electromagnet to substantially zero. This unlinks the two generator parts as there is no longer a magnetic interaction.

[0119] Subsequently or alternatively to step 350, the method 300 may include, in the stiction recovery operating mode, step 360, in which: the generator is operated in a or the motoring mode. Reversing the direction of rotation of the generator has the effect of applying a braking effect to both rotors. Once the two rotors are sufficiently slowed, the kinetic energy of the fluid (for example, air) is sufficient to restart each rotor in its intended direction of rotation, and thus for the stiction condition to be resolved.

[0120] Once the stiction condition is resolved, the method 300 may return to step 310.

[0121] As discussed, the present invention provides a way to maximise the power output from the turbine 100, 200 whilst achieving a steady state of operation and having a shared generator 130, 230 for two rotors. A suitable optimisation strategy for this purpose will now be described. Considering first the input of power to the system: the power input to the turbine 100, 200 from the fluid (for example, wind) is defined in Equation 1 :

[0122] P = )1T1“er0+ )2T^er° (1)

[0123] In Equation 1 : P is the power input to the turbine 100, 200 from the fluid; a>1is the rotational speed of the first rotor 110, 210; a>2is the rotational speed of the second rotor 120, 220;Taero forqueexerted by the fluid on the first rotor 110, 210 (that is, the aerodynamic torque on the first rotor 110, 210); and2erois the torque exerted by the fluid on the second rotor 120, 220 (that is, the aerodynamic torque on the second rotor 120, 220).

[0124] As discussed, the optimisation strategy seeks to make the aerodynamic torques exerted on the first 110 and second 120 equal, as close to equal as is practically feasible, and / or to generally minimise the difference between the aerodynamic torques, for example such that the difference falls within a threshold amount. This is to encourage steady state operation, whilst seeking maximum power output (rather than simply controlling the turbine 100, 200 to have a generator torque which only seeks to maximise the power produced). To achieve this goal, the optimisation strategy includes a penalty function which is a monotonically increasing (that is, can only increase) function of the relative difference between the two aerodynamic torques. This is defined in Equation 2:

[0125] R = f (r“ero- Tero)2) (2)

[0126] In Equation 2: R is the penalty term; t°-erois the aerodynamic torque on the first rotor 110, 210, as before; and2erois the aerodynamic torque on the second rotor 120, 220, as before.

[0127] In alternative arrangements, f may also depend upon quantities which describe the fluid conditions experienced by the turbine 100, 200. For example, if the fluid is very turbulent, then a more cautious approach should be taken, more strictly keeping the aerodynamic torques equal. This is because, in these situations (of turbulent fluid), rotation of the first 110, 210 and second 120, 220 rotors in the same direction - that is, corotation - is more likely to arise. In such conditions, the turbine 100, 200 should thus be carefully controlled. For completeness, corotation is undesirable because this results in low to no power output from the turbine 100, 200.

[0128] The aim of the optimisation is to maximise the difference between the power input to the turbine 100 from the fluid (that is, P) and the penalty term (that is, R). This is defined mathematically in a target function, Equation 3:

[0129] T = P - R (3)

[0130] Thus, put simply, the aim of the optimisation is to maximise the target T.

[0131] For completeness, it is advantageous to calculate the target function for the specific location in space that the turbine 100, 200 is deployed in. This is because fluid conditions differ between different locations, which can alter the target function. For example, in the case of a wind turbine: in urban environments, the wind flow can be very disturbed and non-uniform; whereas in rural environments, the wind flow can be very uniform.

[0132] Considering next the generator 130, 230, the voltage across the generator 130, 230 is proportional to its rotational speed, according to Equation 4:

[0133] V = Ku (4)

[0134] In Equation 4: V is voltage across the generator; K is a constant (described in more detail below in relation to Equation 5); and co is the rotational speed of the generator.

[0135] For simplicity of presentation, the generator is assumed to be an ideal generator. Thus, in Equation 4, the voltage is presented as proportional to the rotational speed of the generator. However, in practice, this may not strictly be the case. One reason for this is that the coils themselves have a resistance. As a result, some voltage will be dropped across the coils. The proportion of the voltage that is dropped across the coils will depend on the size of the load attached.

[0136] Returning to the example optimisation strategy: the torque acting against the motion of the generator 130, 230 is proportional to the current flowing through the generator, according to Equation 5:

[0137] T9= K l (5)

[0138] In Equation 5: is the generator torque; K is a constant dependent upon the specific configuration of the generator and which can be varied by varying the magnetic field of the electromagnet (that is, by varying the current in the electromagnet(s)); and I is the current flowing through the generator.

[0139] The generator torque is then defined using Equation 6:

[0140] In Equation 6: is the generator torque, as before; K is a constant, as before; co is the rotational speed of the generator, as before; R is the effective load on the generator; and A is introduced for simplicity. A can be varied by varying the magnetic field of the electromagnet and / or by varying the effective load on the generator. As noted, the effective load on the generator can be varied by varying the current in the electrical winding(s).

[0141] In this example, the generator 130, 230 is a direct drive contrarotating generator. In such arrangements, the rotational speed of the generator is the sum of the rotational speeds of the two rotors. As such, the generator torque may also be defined using Equation 7:

[0142] Tg= ACc i + Ct>2) (7) are all as before. The equation of motion for each rotor is then defined using Equation 8:

[0143] Where i = 1 , 2 (i.e., where i identifies the rotor), I is the moment or inertia, and all other terms are as before. For completeness, terms describing friction may also be included, however - for simplicity of presentation and ease of understanding - these terms are omitted here.

[0144] Accordingly, since the rotational speed of each rotor is (directly or indirectly) measured, and since A is known, it is possible to determine the aerodynamic torques experienced by the first and second rotors at any time.

[0145] The next step in the optimisation method is to optimise the function expressed in Equation 3 by changing A. This is so that, in the next time step, Tt+1is maximised. This can be approximated using Equation 9:

[0146] Tt+1= Tt+ T At(9)

[0147] Where Atis the length of the next time step. Thus T is maximised with respect to A.

[0148] In one example, this is achieved by a gradient ascent method. In such an approach, the gradient of T is calculated, and then A adjusted in that direction. This is shown in Equation 10:

[0149] Where <x is a parameter determining the step size. A small value for <x will lead to a slower response of the turbine and hence a longer time to reach the optimal A value and maximum power. A large value for <x will however increase the risk that the optimal A value is overshot, potentially leading to oscillatory dynamics. The value for <x must thus be carefully chosen.

[0150] Various different mathematical approaches can however be taken to the optimisation, as will occur to the skilled person. For completeness it is noted that in one mathematical approach, such as the gradient ascent method mentioned above, the fluid (for example, wind) condition is assumed to not be changing. This allows for a relatively simple approach which is less burdensome on the processor. However, if the fluid condition changes - for example, there is a gust of wind - then this approach is no longer accurate. For example, in a sudden gust, the aerodynamic torque experienced by one or both rotors would increase regardless of change in rotational speed. This disadvantage can however be managed by limiting the rate of change of A.

[0151] An alternative approach is to use a model in which aerodynamic torque is modelled as a function of the wind condition and rotational speed. This is implementable using a look-up table and interpolation. However, any other function approximation scheme could be used. Initial datapoints to construct the look-up table are obtainable from simulations or from controlled experiments. Further still, alternatively, a machine learning technique, such as a neural network or a random forest, may be used instead to approximate the function. In a machine learning technique, the mentioned initial data points may be the training data. The advantage of using a model in which aerodynamic torque is modelled as a function of the fluid (for example, wind) condition and rotational speed is that the model may be updated over time to reflect one or both of: the operating environment of the turbine; and / or changes in characteristics of the turbine, for example due to sullying of the turbine blades.

[0152] Alternatively, a mathematical optimisation approach may be taken in which both of the above described approaches are employed. That is, the approach in which the fluid is assumed not to have changed is used in parallel with the approach where the fluid condition is considered in the optimisation. If the outputs from the two methods are sufficiently close - for example, within a threshold amount of each other - then it can be assumed that the fluid conditions have not changed. In that case, the average value can be used for the optimisation step output. On the other hand, if the outputs from the two methods are not sufficiently close - for example, not within a threshold amount of each other - then it can be assumed that the fluid conditions have changed. In this case, the model value where the fluid conditions are considered is taken for the optimisation step output. Furthermore, in that case, the influence of the changing fluid condition on the aerodynamic torque can be estimated and that quantity used in the penalty term R (first introduced in Equation 2).

[0153] Turning lastly to Figure 4, Figure 4 shows a schematic and simplified representation of a computing device 400 which can be used to perform the methods described herein, either alone, in combination with other computer devices or apparatuses, or as part of a cloud computing arrangement. The control system previously described may be considered as comprising the computing device 400.

[0154] The computing device 400 comprises various data processing resources such as a processor 402 (in particular a hardware processor) coupled to a central bus structure. Also connected to the bus structure are further data processing resources such as memory 404. Optionally, a display adapter 406 connects a display device 408 to the bus structure. One or more user-input device adapters 410 may also be present, and which connect a user-input device 412, such as a keyboard and / or a mouse to the bus structure. One or more communications adapters 414 may also be connected to the bus structure to provide connections to other computer systems 400 and other networks.

[0155] In operation, the processor 402 of computing device 400 executes a computer program comprising computer-executable instructions that may be stored in memory 404. When executed, the computer-executable instructions may cause the computing device 400 to perform one or more of the methods described herein. The results of the processing performed may optionally be displayed to a user via the display adapter 406 and display device 408. User inputs for controlling the operation of the computing device 400 may be received via the user-input device adapters 410 from the user-input devices 412.

[0156] It will be apparent that some features of computing device 400 shown in Figure 4 may be absent in certain cases. For example, one or more of the plurality of computing devices 400 may have no need for display adapter 406 or display device 408. This may be the case, for example, for particular server-side computing devices 400 which are used only for their processing capabilities and do not need to display information to users. Similarly, user input device adapter 410 and user input device 412 may not be required. In its simplest form, computing device 400 comprises processor 402 and memory 404.

[0157] While various specific combinations of components and method steps have been described, these are merely examples. Components and method steps may be combined in any suitable arrangement or combination. Components and method steps may also be omitted to leave any suitable combination of components or method steps.

[0158] The described methods may be implemented using computer executable instructions. A computer program product or computer readable medium may comprise or store the computer executable instructions. The computer program product or computer readable medium may comprise a hard disk drive, a flash memory, a read-only memory (ROM), a CD, a DVD, a cache, a random-access memory (RAM) and / or any other storage media in which information is stored for any duration (e.g., for extended time periods, permanently, brief instances, for temporarily buffering, and / or for caching of the information). A computer program may comprise the computer executable instructions. The computer readable medium may be a tangible or non-transitory computer readable medium. The term “computer readable” encompasses “machine readable”.

[0159] The singular terms “a” and “an” should not be taken to mean “one and only one”. Rather, they should be taken to mean “at least one” or “one or more” unless stated otherwise. The word “comprising” and its derivatives including “comprises” and “comprise” include each of the stated features, but does not exclude the inclusion of one or more further features.

[0160] The above implementations have been described by way of example only, and the described implementations are to be considered in all respects only as illustrative and not restrictive. It will be appreciated that variations of the described implementations may be made without departing from the scope of the disclosure. It will also be apparent that there are many variations that have not been described, but that fall within the scope of the appended claims.

Claims

CLAIMS1 . A method of operating a turbine, the turbine comprising: a first rotor arranged to rotate about a rotor axis; a second rotor arranged to rotate about the rotor axis, the first and second rotors arranged for contrarotation about the rotor axis; a generator comprising first and second generator parts, wherein: the first generator part comprises at least one electromagnet and is arranged to rotate with the first rotor about the rotor axis; and the second generator part comprises at least one electrical winding and is arranged to rotate with the second rotor about the rotor axis; wherein the method comprises: determining a first parameter, wherein the first parameter is the rotational speed of one of the first rotor or the second rotor; determining a second parameter, wherein the second parameter is the rotational speed of the other of the first rotor or the second rotor; and based on the determined first and second parameters, controlling the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

2. The method of claim 1 , wherein the step of determining the second parameter comprises: directly measuring the rotational speed of the other of the first or second rotor; or: determining the voltage in the at least one electrical winding; determining the first parameter; and determining the second parameter based on the determined voltage and the determined first parameter.

3. The method of claim 1 or claim 2, wherein the first parameter is determined by directly measuring the rotational speed of the one of the first or second rotor.

4. The method of any preceding claim, wherein the method further comprises: detecting that the difference between the aerodynamic torque acting on each of the first and second rotors is above a threshold; and in response to detecting that the difference is above the threshold, entering a stiction recovery operating mode.

255. The method of claim 4, wherein the method further comprises, in the stiction recovery operating mode: decoupling the first and second generator parts by reducing the current in the at least one electromagnet to substantially zero.

6. The method of claim 4 or claim 5, wherein the method further comprises, in the stiction recovery operating mode: operating the generator in a motoring mode.

7. A turbine comprising: a first rotor arranged to rotate about a rotor axis; a second rotor arranged to rotate about the rotor axis, the first and second rotors arranged for contrarotation about the rotor axis; a generator comprising first and second generator parts, wherein: the first generator part comprises at least one electromagnet and is arranged to rotate with the first rotor about the rotor axis; and the second generator part comprises at least one electrical winding and is arranged to rotate with the second rotor about the rotor axis; measuring means for measuring a first parameter, wherein the first parameter is the rotational speed of one of the first rotor or the second rotor; the or further measuring means for measuring a second parameter, wherein the second parameter is: the rotational speed of the other of the first rotor or the second rotor; or the voltage in the at least one electrical winding; and a control system configured to: determine the rotational speed of the one of the first or second rotor based on the first parameter; determine the rotational speed of the other of the first or second rotor based on the second parameter; and control, based on the determined rotational speeds, the current in the at least one electromagnet and / or the at least one electrical winding such that the difference between the aerodynamic torque acting on each of the first and second rotors is minimised.

8. The turbine of claim 7, wherein the measuring means directly measures the first parameter.

9. The turbine of claim 7 or claim 8, wherein: the second parameter is the rotational speed of the other of the first rotor or the second rotor; and the or the further measuring means directly measure the second parameter.

10. The turbine of claim 7 or claim 8, wherein: the second parameter is the voltage in the at least one electrical winding; and the control system is configured to determine the rotational speed of the other of the first or second rotor based on the second parameter and the determined rotational speed of the one of the first or second rotor.

11. The turbine of any of claims 7 to 10, wherein the control system is configured to: monitor the difference between the aerodynamic torque acting on each of the first and second rotors; and in response to detecting that the difference is above a threshold, enter a stiction recovery operating mode.

12. The turbine of any of claims 7 to 11 , wherein the control system is configured to: detect that the first and second rotors are rotating in the same direction about the rotor axis; and in response to detecting that the first and second rotors are rotating in the same direction, enter a or the stiction recovery operating mode.

13. The turbine of claim 11 or claim 12, wherein the control system is configured to, when in the stiction recovery operating mode: decouple the first and second generator parts by reducing the current in the at least one electromagnet to substantially zero; and / or operate the generator in a motoring mode.

14. The turbine of any of claims 7 to 13, wherein the turbine is a vertical axis wind turbine ‘VAWT’ or a horizontal axis wind turbine ‘HAWT’.

15. The turbine of any of claims 7 to 14, wherein the turbine is: a wind turbine and the fluid is air; or a propeller and the working fluid is a liquid.

Images