Methods for detecting structural damage of a wind turbine blade and wind turbines

Abstract

The present disclosure relates to a method (100) for detecting a structural damage of a first blade (22) of a wind turbine (10). The wind turbine (10) comprises a rotor (18) with one or more blades (22). The method (100) comprises measuring time series of data indicative of loads on the first blade (22) including at least a first time series of data measured with a blade sensor (991, 992) over a measurement period. The method (100) also comprises calculating a first statistical characteristic from the measured first time series of data and detecting the structural damage of the first blade (22) at least partially based on the first calculated statistical characteristic. In a further aspect, the present disclosure relates to a wind turbine (10) comprising a control unit configured to carry out such method (100).

Description

GENERAL ELECTRIC RE OVABLES ESPANA S.L. AUGUST 29, 2024701187-WO-1 P5526PC00METHODS FOR DETECTING STRUCTURAL DAMAGE OF A WIND TURBINE BLADE AND WIND TURBINES

[0001] The present disclosure relates to wind turbines, and more particularly, to methods for detecting structural damage of wind turbine blades and to wind turbines incorporating such methods.BACKGROUND

[0002] Modern wind turbines are commonly used to supply electricity into the electrical grid. Wind turbines of this kind generally comprise a tower and a rotor arranged on the tower. The rotor, which typically comprises a hub and a plurality of blades, is set into rotation under the influence of the wind on the blades. Said rotation generates a moment that is normally transmitted through a rotor shaft to a generator, either directly (“directly driven” or “gearless”) or through the use of a gearbox. This way, the generator produces electricity which can be supplied to the electrical grid.

[0003] The wind turbine hub may be rotatably coupled to a front of the nacelle. The wind turbine hub may be connected to a rotor shaft, and the rotor shaft may then be rotatably mounted in the nacelle using one or more rotor shaft bearings arranged in a frame inside the nacelle. The nacelle is a housing arranged on top of a wind turbine tower that may contain and protect the gearbox (if present) and the generator (if not placed outside the nacelle) and, depending on the wind turbine, further components such as a power converter, and auxiliary systems.

[0004] Wind turbine blades extract kinetic energy from the wind and transform it into rotational kinetic energy. In order to extract more energy from the wind, the length of the blades has steadily increased in modern wind turbines. As a result, higher physical loads are introduced into the blade and related components. Blades need to remain structurally efficient to withstand all loads while remaining as light as possible to facilitate logistics and installation. Specifically, both extreme and fatigue loads need to be withstood while minimizing weight and cost of the blades.

[0005] Wind turbine blades may deteriorate or get damaged during the lifetime of a wind turbine. In particular, structural damage of the blades can result in a reduced performance ofthe wind turbine. Even more importantly, structural damage can lead to catastrophic failure, which can result in a collapse of the damaged blade or even of the complete wind turbine. Such catastrophic failures derive in safety risks and significant costs. In particular, the repair and / or replacement of a damaged blade has a very high direct cost and it also requires prolonged downtime of the wind turbine, thus reducing the annual energy yield of the affected wind turbine.

[0006] Such effects are exacerbated when considering offshore wind turbines, i.e. wind turbine installed in the sea. In such a case, operations are more difficult and associated costs escalate.

[0007] Different factors, either during manufacturing, installation or operation, can lead to an eventual structural damage of a wind turbine blade. Manufacturing defects or inconsistencies can result in the presence of voids, delamination, or non-uniform distribution of resin in the wind turbine blade. Such defects can result in increased vulnerability of the blade and in premature structural failures after repeated load cycles due to, e.g. the formation of micro-cracks. In some cases, an improper design, e.g. inadequate materials thickness or inadequate material selection, can also lead to premature weakening of the blade structure.

[0008] During operation of the wind turbine, environmental conditions can also result in structural defects in the blades. Blades are exposed to lightning, UV radiation or humid environments and particularly in the case of offshore wind turbines salt particles in the wind. In some cases, such environmental conditions can induce weakening or localized damage, which can eventually result in a structural damage of the blade. Besides, damage induced by impacts with foreign objects, e.g. birds, or hail, can also induce localized damage which can extend and jeopardize the structural health of the blade. In particular, leading edge erosion, resulting from repeated impacts with rain, dust or airborne particles, can lead to reduced structural integrity.

[0009] Furthermore, wind forces and vibrations can occur during operation which are higher than expected during the design phase of the wind turbine. Such forces and vibrations may weaken, or even break, a wind turbine blade. In particular, specific wind farms can exhibit wind with more turbulence than expected, or more storms than expected may occur. Other conditions, such as excessively high or excessively low temperatures and humidity, may also deteriorate the structure of a wind turbine blade.

[0010] In order to mitigate the impact of such blade damage events, known methods include periodic blade inspections by maintenance operators. Visual inspection and analysis with dedicated equipment can be carried out by maintenance personnel. Although suchinspections can effectively identify deteriorated portions and blade defects, they also exhibit some drawbacks. In particular, inspecting an inside of a wind turbine blade may be difficult and time consuming for operators. Also, varying weather conditions and wind turbines installed in places which are difficult to access, e.g. offshore, may pose a challenge for safely reaching and accessing an inside of the wind turbine blades. Furthermore, such inspections can only be conducted while the wind turbine is not operating. Accordingly, either the overall downtime of the wind turbine needs to be increased to enable high frequency inspections, which obviously reduces energy yield, or the frequency of the inspections is kept relatively low, thus increasing the risk of potential blade damages occurring in between inspections.

[0011] Other known methods comprise the use of relatively expensive image acquisition systems, e.g. cameras, arranged in the wind turbine blades or on the nacelle. Such cameras allow an online inspection of the blade, which provides a safer and faster process than regular inspections by maintenance personnel. Nevertheless, the use of such image acquisition systems is relatively expensive. Besides, the cameras themselves may need replacement due to wear and tear, and repairing or replacing them can also be costly. Furthermore, the ability to detect blade damage is constrained by the specific arrangement and by the number of cameras. In particular, early detection of blade damage, which is desired to avoid potentially catastrophic consequences, may be particularly challenging with such systems.

[0012] Accordingly, improved methods for the detection of structural defects in wind turbine blades would be desired. In particular, it is desired that such detections can be carried out without stopping the wind turbine, in a cost-effective manner, and with enough sensitivity to detect structural damage at a relatively early stage such that catastrophic failures can be prevented.

[0013] The present disclosure aims at providing methods and systems to address at least one of the above-mentioned aspects by providing improved detection of structural defects in wind turbine blades.SUMMARY

[0014] In an aspect of the present disclosure, a method for detecting structural damage of a first blade of a wind turbine is provided. The wind turbine comprises a rotor with a plurality of blades. The method comprises measuring time series of data indicative of loads on the first blade including at least a first time series of data measured with a blade sensor over a measurement period. The method further comprises calculating a first statistical characteristicfrom the measured first time series of data, and detecting the structural damage of the first blade. The detection is at least partially based on the first calculated statistical characteristic.

[0015] According to this aspect of the disclosure, an early detection of a blade structural damage is achieved. In particular, structural damage can be detected before significant, and even catastrophic, failure occurs.

[0016] Inventors have realized that damaged wind turbine blades exhibit a different load distribution and behavior than “healthy” blades. Accordingly, data indicative of blade loads can be used as a proxy for blade structural damage detection.

[0017] The structural change in the blades is dependent on the failure mechanism, and it can result in measurements that are either attenuated or amplified with respect to a healthy blade. That is, depending on the type and arrangement of the blade sensor, and depending on the failure mechanism, abnormal values can be detected by the blade sensors. Hence, degradation over relatively short periods of time, e.g. in the range of few hours, may be detected on the basis of the data obtained with the blade sensors along such period of time.

[0018] In examples, a blade comprising multiple blade sensors may be envisaged. The different blade sensors may be indicative of different loads on the blade. Time series of data may be measured for each or multiple of the sensors and a corresponding statistical characteristic may be derived from each such time series of data. Depending on the location of the failure or damage, a first sensor may provide an attenuated reading with respect to a healthy wind turbine whereas a second sensor may provide an amplified reading. Such abnormal readings may be represented by the corresponding statistical characteristics. Moreover, the blade defect may progress with load cycles, i.e. during wind turbine operation. Consequently, measured data may also vary over time, so that it may further attenuate or amplify the corresponding measured signals.

[0019] In order to properly account for the dynamics of the system, a statistical processing of the measured data is carried out. In this manner, the output from the blade sensors are not directly considered, but a derived statistical characteristic, e.g. average, median, standard deviation, variance, etc. is employed.

[0020] In some examples, the blade sensor or sensors may comprise a dedicated sensor or a suite of dedicated sensors, which may be arranged on the blade in order to implement the method according to this aspect of the disclosure. In other examples, already existing blade sensors may be employed. In particular, some wind turbines may comprise blade load sensors, which may be used for control purposes, e.g. for blade pitch control or for ice detection. In suchcases, data obtained from such existing blade load sensors may be used to implement the present method to detect structural damage of the blades.

[0021] Throughout the present disclosure, data indicative of loads is to be understood as any data, obtained with a blade sensor, that is affected by prevailing loads. In particular, such data may be obtained with a load sensor, with a sensor measuring a deformation, or with any other type of sensor from which loads can be inferred. Accordingly, measured data may be obtained and statistically processed in the form of a load, e.g. as a moment or as a force, but also in the form of other magnitudes such as, but not limited to, a deformation, a displacement, an acceleration, an electrical resistance, etc.

[0022] In another aspect of the disclosure, a method of operating a wind turbine is provided. The wind turbine comprises a rotor with a plurality of blades. The method comprises measuring time series of data indicative of loads on a first blade including at least a first time series of data measured with a blade sensor over a measurement period. The method also comprises calculating a first statistical characteristic from the measured first time series of data, and detecting a structural damage of the first blade at least partially based on the calculated first statistical characteristic. Furthermore, the method of operating a wind turbine also comprises stopping or shutting down the wind turbine upon detection of the structural damage.

[0023] According to this aspect of the disclosure, a safer operation of a wind turbine is achieved. Hence, the method allows an early detection of a structural damage in a blade. Depending on the failure mechanism, structural damage can progress very rapidly. Accordingly, in order to prevent a catastrophic event, e.g. breaking of the blade or even collapse of the wind turbine, the method according to the present aspect comprises discontinuing operation of the wind turbine.

[0024] By stopping the wind turbine before a catastrophic failure occurs, an opportunity to repair and / or replace the structurally damaged blade is provided. Such operations are more effective, both in terms of cost and time, than those arising after a blade is broken. Furthermore, the detected structural damage may be used to trigger a controlled shut down of the wind turbine, thus ensuring the structural integrity of the wind turbine.

[0025] Still in a further aspect of the disclosure, a wind turbine is provided. The wind turbine comprises a tower, a rotor with one or more blades, and a control unit. The control unit is configured for receiving a first time series of data indicative of a load on a first blade over a measurement period, and for calculating a first statistical characteristic from the received first series of data. The control unit is also configured for detecting structural damage of the first blade at least partially based on the first calculated statistical characteristic.

[0026] According to this further aspect of the disclosure, a wind turbine with improved reliability is obtained. In particular, the wind turbine according to this aspect of the disclosure takes advantage of the previously described aspects by adopting a method that allows early detection of blade structural damage.BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Non-limiting examples of the present disclosure will be described in the following, with reference to the drawings, in which:Figure 1 illustrates a perspective view of one example of a wind turbine;Figure 2 illustrates a simplified, internal view of one example of the nacelle of the wind turbine of the figure 1 ;Figures 3A - 3B illustrate a perspective view (3A) and a cross-section view (3B) of a wind turbine blade with blade load sensors according to an example;Figure 4 shows a flowchart of an example of a method for detecting structural damage of a wind turbine blade;Figure 5 schematically illustrates a block diagram of an example of a method for detecting structural damage of a wind turbine blade;Figure 6 schematically illustrates a simulation of the results obtained with a method for detecting structural damage of a wind turbine blade according to an example;Figure 7 schematically illustrates the results obtained with a method for detecting structural damage of a wind turbine blade according to another example;Figure 8 shows a flowchart of an example of a method for operating a wind turbine; andFigure 9 schematically illustrates a layout of components in a wind turbine with a control unit configured for detecting blade damage according to an example.DETAILED DESCRIPTION OF EXAMPLES

[0028] Reference now will be made in detail to embodiments, one or more examples of which are illustrated in the drawings. Each example is provided by way of explanation only, not as a limitation. In fact, it will be apparent to those skilled in the art that various modifications and variations can be made in the present disclosure. For instance, features illustrated or described as part of one example can be used with another example to yield a still furtherexample. Thus, it is intended that the present disclosure covers such modifications and variations as come within the scope of the appended claims and their equivalents.

[0029] Figure 1 is a perspective view of an example of a wind turbine 10. In the example, the wind turbine 10 is a horizontal-axis wind turbine. Alternatively, the wind turbine 10 may be a vertical-axis wind turbine. In the example, the wind turbine 10 includes a tower 15 that extends from a support system 14 on a ground 12, a nacelle 16 mounted on tower 15, and a rotor 18 that is coupled to nacelle 16. The rotor 18 includes a rotatable hub 20 and at least one rotor blade 22 coupled to and extending outward from the hub 20. In the example, the rotor 18 has three rotor blades 22. In an alternative embodiment, the rotor 18 includes more or less than three rotor blades 22. The tower 15 may be fabricated from tubular steel to define a cavity (not shown in Figure 1) between a support system 14 and the nacelle 16. In an alternative embodiment, the tower 15 is any suitable type of a tower having any suitable height. According to an alternative, the tower can be a hybrid tower comprising a portion made of concrete and a tubular steel portion. Also, the tower can be a partial or full lattice tower.

[0030] The rotor blades 22 are spaced about the hub 20 to facilitate rotating the rotor 18 to enable kinetic energy to be transferred from the wind into usable mechanical energy, and subsequently, electrical energy. The rotor blades 22 are mated to the hub 20 by coupling a blade root region 24 to the hub 20 at a plurality of load transfer regions 26. The load transfer regions 26 may have a hub load transfer region and a blade load transfer region (both not shown in figure 1). Loads induced to the rotor blades 22 are transferred to the hub 20 via the load transfer regions 26.

[0031] In examples, the rotor blades 22 may have a length ranging from about 15 meters (m) to about 100 m or more. Rotor blades 22 may have any suitable length that enables the wind turbine 10 to function as described herein. For example, non-limiting examples of blade lengths include 20 m or less, 37 m, 48.7 m, 50.2m, 52.2 m or a length that is greater than 91 m. As wind strikes the rotor blades 22 from a wind direction 28, the rotor 18 is rotated about a rotor axis 30. As the rotor blades 22 are rotated and subjected to centrifugal forces, the rotor blades 22 are also subjected to various forces and moments. As such, the rotor blades 22 may deflect and / or rotate from a neutral, or non-deflected, position to a deflected position.

[0032] Moreover, a pitch angle of the rotor blades 22, i.e., an angle that determines an orientation of the rotor blades 22 with respect to the wind direction, may be changed by a pitch system 32 to control the load and power output by the wind turbine 10 by adjusting an angular position of at least one rotor blade 22 relative to wind vectors. Pitch axes 34 of rotor blades 22 are shown. During operation of the wind turbine 10, the pitch system 32 may particularly change a pitch angle of the rotor blades 22 such that the angle of attack of (portions of) therotor blades are reduced, which facilitates reducing a rotational speed and / or facilitates a stall of the rotor 18.

[0033] In the example, a pitch angle of each rotor blade 22 is controlled individually by a wind turbine controller 36 or by a pitch control system 80 (see Figure 2). Alternatively, the blade pitch for all rotor blades 22 may be controlled simultaneously by said control systems.

[0034] Further, in the example, as the wind direction 28 changes, a nacelle 16 may be rotated about the longitudinal axis of the tower, i.e. about a yaw axis 38 to position the rotor blades 22 with respect to wind direction 28.

[0035] In the example, the wind turbine controller 36 is shown as being centralized within the nacelle 16. However, the wind turbine controller 36 may be a distributed system throughout the wind turbine 10, on the support system 14, within a wind farm, and / or at a remote-control center. The wind turbine controller 36 may include a processor 40 configured to perform some of the methods and / or steps described herein. Further, many of the other components described herein include a processor.

[0036] As used herein, the term “processor” is not limited to integrated circuits referred to in the art as a computer, but broadly refers to a controller, a microcontroller, a microcomputer, a programmable logic controller (PLC), an application specific, integrated circuit, and other programmable circuits, and these terms are used interchangeably herein. It should be understood that a processor and / or a control system can also include memory, input channels, and / or output channels.

[0037] A control system 36 may also include a memory, e.g. one or more memory devices. A memory may comprise memory element(s) including, but not limited to, a computer readable medium (e.g., random access memory (RAM)), a computer readable non-volatile medium (e.g., a flash memory), a floppy disk, a compact disc-read only memory (CD-ROM), a magnetooptical disk (MOD), a digital versatile disc (DVD) and / or other suitable memory elements. Such memory device(s) may generally be configured to store suitable computer-readable instructions that, when implemented by the processor(s) 40, configure the controller 36 to perform, or trigger the performance of, various steps disclosed herein. A memory may also be configured to store data, e.g. from measurements and / or calculations.

[0038] Figure 2 is an enlarged sectional view of a portion of the wind turbine 10. In the example, the wind turbine 10 includes the nacelle 16 and the rotor 18 that is rotatably coupled to the nacelle 16. More specifically, the hub 20 of the rotor 18 is rotatably coupled to an electric generator 42 positioned within the nacelle 16 by the main shaft 44, a gearbox 46, a high-speed shaft 48, and a coupling 50. In the example, the main shaft 44 is disposed at least partiallycoaxial to a longitudinal axis (not shown) of the nacelle 16. A rotation of the main shaft 44 drives the gearbox 46 that subsequently drives the high-speed shaft 48 by translating the relatively slow rotational movement of the rotor 18 and of the main shaft 44 into a relatively fast rotational movement of the high-speed shaft 48. The latter is connected to the generator 42 for generating electrical energy with the help of a coupling 50. Furthermore, a transformer 90 and / or suitable electronics, switches, and / or inverters may be arranged in the nacelle 16 in order to transform electrical energy generated by the generator 42 having a voltage between 400V to 1000 V into electrical energy having medium voltage (e.g. 10 - 35 KV). Said electrical energy is conducted via power cables from the nacelle 16 into the tower 15.

[0039] The gearbox 46, generator 42 and transformer 90 may be supported by a main support structure frame of the nacelle 16, optionally embodied as a main frame 52. The gearbox 46 may include a gearbox housing that is connected to the main frame 52 by one or more torque arms 103. In the example, the nacelle 16 also includes a main forward support bearing 60 and a main aft support bearing 62. Furthermore, the generator 42 can be mounted to the main frame 52 by decoupling support means 54, in particular in order to prevent vibrations of the generator 42 to be introduced into the main frame 52 and thereby causing a noise emission source.

[0040] Optionally, the main frame 52 is configured to carry the entire load caused by the weight of the rotor 18 and components of the nacelle 16 and by the wind and rotational loads, and furthermore, to introduce these loads into the tower 15 of the wind turbine 10. The rotor shaft 44, generator 42, gearbox 46, high-speed shaft 48, coupling 50, and any associated fastening, support, and / or securing device including, but not limited to, support 52, and forward support bearing 60 and aft support bearing 62, are sometimes referred to as a drive train 64.

[0041] In some examples, the wind turbine may be a direct drive wind turbine without gearbox 46. Generator 42 operate at the same rotational speed as the rotor 18 in direct drive wind turbines. They therefore generally have a much larger diameter than generators used in wind turbines having a gearbox 46 for providing a similar amount of power than a wind turbine with a gearbox.

[0042] The nacelle 16 also may include a yaw system which comprises a yaw bearing (not visible in figure 2) having two bearing components configured to rotate with respect to the other. The tower 15 is coupled to one of the bearing components and the bedplate or support frame 52 of the nacelle 16 is coupled to the other bearing component.

[0043] The yaw system may comprise an annular gear and a yaw drive mechanism 56 that may be used to rotate the nacelle 16 and thereby also the rotor 18 about the longitudinal axisof the tower, i.e. about a yaw axis 38 to control the perspective of the rotor blades 22 with respect to the wind direction 28.

[0044] The yaw drive mechanism 56 may comprise a plurality of yaw drives with a motor, a gearbox and a pinion for meshing with the annular gear for rotating one of the bearing components with respect to the other. The annular gear may comprise a plurality of teeth which engage with the teeth of the pinion. In the example of Figure 2, the yaw drives and the annular gear are placed outside the external diameter of the tower. The teeth of the annular gear are outwardly orientated, but in other examples, the annular gear and yaw drives may be arranged at the inside of the tower.

[0045] In some examples, one of the yaw drives may be a “master”, and the other drives may be “slaves” following the instructions of the master or adapting their operation to adapt to the master drive.

[0046] The turbine controller 36 may be communicatively coupled to the yaw drive mechanism 56 of the wind turbine 10 for controlling and / or altering the yaw direction of the nacelle 16 relative to the wind direction 28. As the direction of the wind 28 changes, the wind turbine controller 36 may be configured to control a yaw angle of the nacelle 16 about the longitudinal axis of the tower or yaw axis 38 to position the rotor blades 22, and therefore the rotor 18, with respect to the direction 28 of the wind, thereby controlling the loads acting on the wind turbine 10. For example, the turbine controller 36 may be configured to transmit control signals or commands to the yaw drive mechanism 56 of the wind turbine 10, via a yaw controller or direct transmission, such that the nacelle 16 may be rotated about the longitudinal axis of the tower or yaw axis 38 via a yaw bearing.

[0047] For positioning the nacelle 16 appropriately with respect to the wind direction 28, the nacelle 16 may also include at least one meteorological measurement system which may include a wind vane and anemometer. The meteorological measurement system 58 can provide information to the wind turbine controller 36 that may include wind direction 28 and / or wind speed.

[0048] In the example, the pitch system 32 is at least partially arranged as a pitch assembly 66 in the hub 20. The pitch assembly 66 includes one or more pitch drive systems 68 and at least one sensor 70. Each pitch drive system 68 is coupled to a respective rotor blade 22 (shown in Figure 1) for modulating the pitch angel of a rotor blade 22 along the pitch axis 34. Only one of three pitch drive systems 68 is shown in figure 2.

[0049] In the example, the pitch assembly 66 includes at least one pitch bearing 72 coupled to hub 20 and to a respective rotor blade 22 (shown in Figure 1) for rotating the respective rotorblade 22 about the pitch axis 34. The pitch drive system 68 includes a pitch drive motor 74, a pitch drive gearbox 76, and a pitch drive pinion 78. The pitch drive motor 74 is coupled to the pitch drive gearbox 76 such that the pitch drive motor 74 imparts mechanical force to the pitch drive gearbox 76. The pitch drive gearbox 76 is coupled to the pitch drive pinion 78 such that the pitch drive pinion 78 is rotated by the pitch drive gearbox 76. The pitch bearing 72 is coupled to pitch drive pinion 78 such that the rotation of the pitch drive pinion 78 causes a rotation of the pitch bearing 72.

[0050] Pitch drive system 68 is coupled to the wind turbine controller 36 for adjusting the pitch angle of a rotor blade 22 upon receipt of one or more signals from the wind turbine controller 36. In the example, the pitch drive motor 74 is any suitable motor driven by electrical power and / or a hydraulic system that enables pitch assembly 66 to function as described herein. Alternatively, the pitch assembly 66 may include any suitable structure, configuration, arrangement, and / or components such as, but not limited to, hydraulic cylinders, springs, and / or servomechanisms. In certain embodiments, the pitch drive motor 74 is driven by energy extracted from a rotational inertia of hub 20 and / or a stored energy source (not shown) that supplies energy to components of the wind turbine 10.

[0051] The pitch assembly 66 may also include one or more pitch control systems 80 for controlling the pitch drive system 68 according to control signals from the wind turbine controller 36, especially in case of specific prioritized situations and / or during rotor 18 overspeed. In the example, the pitch assembly 66 includes at least one pitch control system 80 communicatively coupled to a respective pitch drive system 68 for controlling pitch drive system 68 independently from the wind turbine controller 36. In the example, the pitch control system 80 is coupled to the pitch drive system 68 and to a sensor 70. During normal operation of the wind turbine 10, the wind turbine controller 36 may control the pitch drive system 68 to adjust a pitch angle of rotor blades 22.

[0052] According to an embodiment, a power generator 84, for example comprising a battery and electric capacitors, is arranged at or within the hub 20 and is coupled to the sensor 70, the pitch control system 80, and to the pitch drive system 68 to provide a source of power to these components. In the example, the power generator 84 provides a continuing source of power to the pitch assembly 66 during operation of the wind turbine 10. In an alternative embodiment, power generator 84 provides power to the pitch assembly 66 only during an electrical power loss event of the wind turbine 10. The electrical power loss event may include power grid loss or dip, malfunctioning of an electrical system of the wind turbine 10, and / or failure of the wind turbine controller 36. During the electrical power loss event, the powergenerator 84 operates to provide electrical power to the pitch assembly 66 such that pitch assembly 66 can operate during the electrical power loss event.

[0053] In the example, the pitch drive system 68, the sensor 70, the pitch control system 80, cables, and the power generator 84 are each positioned in a cavity 86 defined by an inner surface 88 of hub 20. In an alternative embodiment, said components are positioned with respect to an outer surface of hub 20 and may be coupled, directly or indirectly, to the outer surface.

[0054] Figure 3A shows a schematic view of an example of a wind turbine blade 22. The wind turbine blade 22 has a blade root end 171 and a tip end 151 and comprises a root region 301 closest to the hub 20, a profiled or an airfoil region 341 furthest away from the hub 20 and a transition region 321 between the root region 301 and the airfoil region 341. The blade 22 comprises a leading edge 181 facing the direction of rotation of the blade 22, when the blade is mounted on the hub 20, and a trailing edge 201 facing the opposite direction of the leading edge 181.

[0055] The airfoil region 341 (also called the profiled region) has an ideal or almost ideal blade shape with respect to generating lift, whereas the root region 301 due to structural considerations has a substantially circular or elliptical cross-section, which for instance makes it easier and safer to mount the blade 22 to the hub 20. The airfoil region 341 has an airfoil profile with a chord extending between the leading edge 181 and the trailing edge 201 of the blade 22.

[0056] Figure 3B is a schematic diagram illustrating a cross sectional view of an example of a wind turbine blade 22, e.g. a cross-sectional view of the airfoil region of the wind turbine blade 22. The wind turbine blade 22 comprises a leading edge 181 , a trailing edge 201 , a pressure side shell part 241 , a suction side shell part 261 , a first spar cap 741 , and a second spar cap 761. The wind turbine blade 22 comprises a chord line 388 between the leading edge 181 and the trailing edge 201. The wind turbine blade 22 comprises one or more shear webs 721 , such as a leading edge shear web and a trailing edge shear web. The shear webs 721 could alternatively be a spar box with spar sides, such as a trailing edge spar side and a leading edge spar side. The spar caps 741 , 761 , extending in a substantially spanwise direction within the blade 22, may comprise glass fibers, carbon fibers or a hybrid composition of fibers while the rest of the shell parts 241 , 261 may comprise glass fibers. The spar caps 741 , 761 may be part of, i.e. may be integrally formed, or they may be adhered to the respective blade shells. Furthermore, each shell part 241 , 261 may comprise one or more additional spar caps.

[0057] Figure 4 shows a flowchart of a method 100 for detecting structural damage of a first blade 22 of a wind turbine 10. The wind turbine 10 comprises a rotor 18 with one or more blades 22. The method 100 comprises, at block 110, measuring time series of data indicative of loads on the first blade 22 including at least a first time series of data measured with a blade sensor over a measurement period. Block 120 of the method 100 comprises calculating a first statistical characteristic from the measured first time series of data. Then, block 130, comprises detecting the structural damage of the first blade at least partially based on the first calculated statistical characteristic.

[0058] In an example of the disclosure, detecting the structural damage of the first blade in block 130 may comprise determining if a difference between a value of the first statistical characteristic, calculated in block 120, with respect to a predetermined value exceeds a predetermined threshold.

[0059] According to this example, the predetermined value may be calculated on the basis of simulations and / or on the basis of previously acquired experimental results. Furthermore, different predetermined values may be employed depending on the environmental and / or operating conditions of the wind turbine 10. Indeed, the value of the statistical characteristic may be representative of damage in a wind turbine blade 22 but it may also be influenced by operating conditions of the wind turbine 10, e.g. prevailing wind speed. Accordingly, in order to discriminate between different factors, and to use the value of the statistical characteristic as a proxy for a blade damage detection, the influence of such operating conditions may be accounted for by adjusting the value of the predetermined value. In other words, the first statistical characteristic may be compared with a predetermined value determined under equivalent environmental conditions. As mentioned, such a predetermined value may be obtained from simulations or it may correspond to an experimental value obtained for the blade, under the same environmental conditions, at a different time, e.g. shortly after the wind turbine installation.

[0060] Different statistical characteristics can be derived from the time series of measured data. In particular, any statistical characteristic indicative of a degree of variation of a load over the measurement period may be employed. Accordingly, in an example, the first statistical characteristic may be the standard deviation of the time series of measured data over the measurement period. The use of the standard deviation has been found by the inventors to be especially sensitive to detect blade structural damage. Nevertheless, other statistical characteristics, such as the average, median, or variance, may be employed in other examples of the disclosure.

[0061] In an example, the method may comprise calculating a second statistical characteristic from the measured first time series of data used for the calculation of the first statistical characteristic. The second statistical characteristic may be different from the first statistical characteristic. The method may further comprise detecting the structural damage of the first blade based on both the first statistical characteristic and the second statistical characteristic. According to this example, a more reliable detection may be provided. As an example, a standard deviation of the data of the time series may be used as the first statistical characteristic whereas an average may be employed as the second statistical characteristic. By using two different characteristics, a more robust blade detection method may be obtained.

[0062] In still other examples, the method may comprise measuring different time series of data with different sensors arranged in the first blade, and calculating a corresponding statistical characteristic for each, or at least for some, of the time series of data. The different sensors may be configured for being indicative of a same load or for being indicative of different loads. Furthermore, either the same or different statistical characteristics may be calculated from the different time series of data acquired with the different sensors.

[0063] In these examples, the detection of the structural damage of the first blade may be based on a combination of at least some of the calculated statistical characteristics. Consequently, an enhanced method for blade damage detection may be obtained. Particularly, by using different sensors, including sensors configured for different loads, a more solid and steady detection method may be provided. Specifically, different sensors may be provided, with individual sensors being especially configured, e.g. located, dimensioned or designed, for reacting to specific loads associated with specific potential blade structural problems.

[0064] In any of the examples, a plurality of measurements are obtained with the blade sensor(s) with a certain sampling period, e.g. 40 ms, so as to obtain a time series of data. Such series of data covers a certain measurement period. Different measurement periods may be selected in different examples. In particular, the measurement period may extend for multiple rotations of the rotor 18 of the wind turbine 10, so as to filter out the influence of rotation-based loads on the measured signals. As an example, a measurement period in the range of 10 minutes may be selected.

[0065] The method 100 described with reference to Figure 4 can be carried out under different scenarios. Hence, in an example, the method 100 may be such that measuring the time series of data may comprise measuring while the wind turbine is in operation. In other words, the method 100 can be implemented in an online manner while the wind turbine 10 is generating power.

[0066] According to this example, a more efficient method 100 for detecting blade 22 damage may be provided as energy production may continue unaffected by the implementation of the method. Many hitherto known methods for the detection of blade defects may require a change in the operating parameters of the wind turbine 10 and, in some cases, even a complete stopping of the wind turbine 10, e.g. in cases comprising physical inspection by an operator. On the contrary, the method 100 in some examples according to the present disclosure allows a seamless online detection of blade defects.

[0067] Even if Figure 4 illustrates a single cycle of the method 100, it is understood that the method may be executed repetitively (and also for multiple sensors simultaneously) so as to provide an improved monitoring of the health of the wind turbine blades 22. In some examples, execution of the method 100 according to the flowchart of Figure 4 may be carried out according to a predetermined schedule, i.e. with a predetermined frequency. The method 100 may be executed in a continuous manner, such that the health or state of the blade 22 may be constantly assessed. Accordingly, not only blade damage may be detected, but degradation over a period of time may also be analyzed.

[0068] In an example, the method 100 may be executed in a sequential manner and only once for each measurement period. In other words, a time series of data may be acquired during a first measurement period, e.g. for 10 minutes, and a detection of structural damage may be carried out based on such acquired time series. Subsequently, after the first measurement period, a new time series of data may be acquired for a subsequent measurement period and a further detection may be carried out. Accordingly, in an example comprising a measurement period of 10 minutes, this example may provide an output, i.e. a detection, every 10 minutes.

[0069] In another example of the method 100, a moving or sliding window may be used for the measurement period. In this manner, a more continuous analysis of the health of the blade 22 may be obtained. For instance, the method 100 may be executed at each sampling time, i.e. with a periodicity corresponding to the sampling period, e.g. 40 ms. Thus, at each sampling period, the time series may be updated based on the data measured over the last measurement period, e.g. over the last 10 minutes. This time series may then be used to calculate the statistical characteristic in block 120 of the method and to detect structural damage in block 130. According to this example, a detection may be obtained in a substantially continuous manner, i.e. at each sampling period.

[0070] Different blade sensors may be used in different examples of the method. In particular, sensors may comprise one of: a fiber Bragg grating (FBG) sensor, a strain gauge,an accelerometer, or an induction sensor. Overall, any kind of sensor, capable of responding to different loads of the blade 22, may be considered in multiple variants of the disclosure.

[0071] Moreover, as already mentioned, such blade sensors may be configured to respond to different types of loads. In an example of the method 100, measuring data indicative of loads on the first blade in block 110 may comprise measuring data indicative of loads acting in an edgewise direction and / or in a flapwise direction of the first blade.

[0072] Different blade failure mechanisms may manifest themselves differently. For this reason, an abnormal load behavior may be present in different directions of the blade 22. Besides, loads in different directions may be affected by different external loads. Hence, edgewise loads during operation may be particularly dependent on the weight and inertial forces, whereas flapwise loads may be mostly dependent on external aerodynamic loads, e.g. thrust, shear, turbulence, etc. The effects of such external loads on the structural response of a blade 22 exhibiting structural damage will depend on the nature of such damage, i.e. on the blade failure mechanism. Accordingly, different sensors may be particularly adequate to identify certain types of defects.

[0073] In a variant of this example , calculating the first statistical characteristic in block 120 may comprise calculating a statistical characteristic of an edgewise bending moment and / or of a flapwise bending moment. As already mentioned, different blade failures may result in different responses in either the flapwise or the edgewise direction. Accordingly, by analyzing the moments in those directions, a proper detection of blade damage may be achieved. Furthermore, edgewise and flapwise bending moments may already be acquired in a wind turbine for other purposes, e.g. for blade pitch control. Accordingly, by using these magnitudes, no further elements may need to be provided.

[0074] Different approaches may be used in block 130 to detect a structural damage of the first blade 22 based on the statistical characteristic obtained in block 120. In an example of the disclosure, the method may comprise continuously repeating the measurement of time series of data and the calculation of the first statistical characteristic during an observation period. Detecting the structural damage of the first blade 22 in block 130 may then comprise analyzing an evolution of a value of the first statistical characteristic and / or analyzing a rate of change of the value of the first statistical characteristic over the observation period. In other words, detection of the structural damage may depend on the history of the statistical characteristic, i.e. on the evolution of the statistical over operating hours of the wind turbine 10. The observation period may be a relatively short period of time, e.g. in the range of hours.

[0075] According to this example, abnormal values of the statistical characteristic, e.g. the standard deviation of an edge or flap moment, may be used to infer the existence of a structural damage in the blade 22. In particular, in a variant of the method, structural damage of the first blade may be detected when a variation of the value of the first statistical characteristic within the observation period exceeds a first threshold. In other variants, structural damage may be detected when the rate of change of the value of the first statistical characteristic exceeds a second threshold within the observation period. In still another variant, both previously mentioned variants may be combined in a single implementation. Overall, the detection of a trend change in the response of the blade sensors, as identified by the evolution of the statistical characteristics over a period of time, may be used to characterize a blade damage. Hence, by taking into account the trend over the observation period, the severity and dynamics of the defect may also be considered.

[0076] Thresholds may be adjusted such that jumps or variations of a certain size in the value of the first statistical characteristic, i.e. the statistical characteristic considered for the time series of data for the first blade, may be properly captured and associated to a blade structural failure. Similarly, a fast variation on the value of the first statistical characteristic, i.e. a variation exceeding a certain predetermined threshold, may also be an indication of a damaged blade even if the absolute values of the statistical characteristic itself remain relatively normal. The thresholds may be predetermined on the basis of previously conducted simulations and / or experimental results. They may also be dependent on both the type of sensor and measured data, as well as on the considered blade failure mechanism.

[0077] In order to obtain a proper response, an observation period may be selected such that real trends, i.e. not simply noise or short-term variations induced by varying environmental conditions, are captured when analyzing the values of the first statistical characteristic. As an example, the observation period may be at least five hours, specifically at least ten hours.

[0078] The dynamics of the structural defects may be dependent on the originating mechanism, e.g. manufacturing defect, extreme weather, etc. Nevertheless, inventors have found that an observation period in the range of a few hours can be enough to promptly detect early damage in the blade 22. In this manner, a quick reaction can be provided.

[0079] In some examples, a value and / or an evolution of a value of a statistical characteristic obtained in a monitored wind turbine blade may be compared with a value and / or evolution of a value previously obtained for a wind turbine blade exhibiting a certain damage. In this manner, if values and / or trends similar to those observed are found, not only a blade damage may be identified, but the nature of the damage, e.g. the failure mechanism, may also be identified. Consequently, a more optimum reaction may be carried out.

[0080] Most wind turbines 10 exhibit a plurality of blades 22. Individual wind turbine blades 22 are subjected to different loads during the rotation of the rotor due to the influence of, e.g. gravity or wind shear. However, the overall loading of all the blades 22 over sufficiently long periods of time, i.e. periods of time clearly exceeding the rotation period of the wind turbine rotor 18, tend to be substantially uniform. Consequently, in some examples of the disclosure, statistical characteristics obtained for different blades may be compared across blades as a proxy to detect blade structural damage in any of the blades 22 of the wind turbine 10.

[0081] Hence, a method for detecting structural damage in a first blade according to an example may comprise carrying out the blocks shown in Figure 4 for the first blade, but also measuring time series of data indicative of loads on a second blade including at least a first time series measured with a blade sensor over a measurement period. The method may then comprise calculating a second statistical characteristic from the measured first time series of data acquired for the second blade. Finally, the method may comprise detecting the structural damage of the first blade by comparing the first statistical characteristic with the second statistical characteristic.

[0082] In an example, the measurement period used for the measurement on the second blade may be the same as the measurement period used for the measurements on the first blade.

[0083] According to this example, advantage may be taken of the fact that both the first and the second blade 22 may be facing substantially equivalent external loads. Consequently, a deviation in the corresponding statistical characteristics can be used as an indication of damage in one of the blades or, more particularly, in the first blade.

[0084] In a variant of this method, the first statistical characteristic and the second statistical characteristic may correspond to the same statistical property, e.g. both the first and second statistical characteristic may correspond to an average, or both to a standard deviation, or both to a variance.

[0085] According to this variant, a more direct and easier analysis may be provided by using a comparison of the same property. The statistical characteristic may be a standard deviation, which inventors have found to be particularly sensitive. As an example, a standard deviation of a load, e.g. an edge or flap moment, in a first blade may be compared with a standard deviation, i.e. the same statistical characteristic, of the same load, e.g. the edge or flap moment, in the second blade.

[0086] Furthermore, in a variant, a further statistical characteristic or behavior may be used to identify a damaged blade. The mere comparison between a first statistical characteristicobtained in a first blade and a second statistical characteristic obtained in the second blade may allow identification of a damage in one of the two blades when, e.g. a discrepancy between the two statistical characteristics is found. Nevertheless, in order to identify the damaged blade, i.e. in order to distinguish the damaged blade from the “healthy” blade, a further statistical characteristic obtained may be used. Such further statistical characteristic may comprise a statistical characteristic based on a further time series of data acquired in another component of the wind turbine, such as in a third wind turbine blade.

[0087] Similarly to previous examples, i.e. examples not involving comparison across different blades, variants of the examples including such comparison may also comprise a consideration of the historical evolution of the different statistical characteristics. In this manner, trends may be analyzed over certain periods of time and a more informed action may be implemented. In a variant, the method may comprise continuously repeating, during an observation period, the measurement of time series of data for the first and the second blade and the calculation of the corresponding statistical characteristics. The calculated first statistical characteristic may be continuously compared with the second statistical characteristic, and detecting the structural damage of the first blade may comprise analyzing an evolution or trend of the comparison over the observation period.

[0088] According to this variant, a more reliable detection may be provided. In particular, short term effects, which may result in abnormal momentary behavior observed in a specific blade may be ruled out. Furthermore, the analysis of the comparison over a sufficiently long observation period, e.g. exceeding ten hours, may also help to discriminate between potential problems with the blade sensors and real blade structural damage. Specifically, progressive degradation may be characterized by increasing deviations over the observation period, whereas malfunctioning of the blade sensors may be characterized by abrupt changes in the corresponding readings. In other words, instead of simply relying on instantaneous differences between the statistical characteristics, the trends of such differences may be considered to provide an improved response.

[0089] In order to further enhance the detection of blade failures in wind turbines comprising a plurality of blades, the method may comprise measuring time series of data in all the blades over the measurement period, the data being indicative of loads on the corresponding blade. The method may further comprise calculating the corresponding statistical characteristics from the measured time series of data. Accordingly, detecting the structural damage of the first blade may be based on the statistical characteristics calculated for all the blades. Specifically, it may be based on a comparison of the statistical characteristic calculated for the first blade with the statistical characteristics calculated for the other blades.

[0090] Moreover, as in the previous example, time series of data and statistical characteristics may be obtained for all the blades in a repetitive manner so as to take trends into consideration for subsequent actions.

[0091] Such an approach may ensure that blade failures in any of the blades 22 may be properly accounted for. Figure 5 schematically illustrates a block diagram 300 of an example of a method for detecting structural damage in a wind turbine 10 comprising three blades 22. According to this example, a data measuring module 310 may be provided to measure data. The data measuring module 310 may comprise three sub-submodules, 311-313. Each submodule may be configured to acquire data from a respective blade 22.

[0092] Different data may be acquired for each blade 22 in the data measuring module 310, such that different time series of data may be acquired for each blade 22. As an example, sub-modules 311-313 may acquire edgewise and / or flapwise moments for the corresponding blades 22 at a certain longitudinal position of the blade 22. In this manner, different time series of data, each corresponding to an edgewise moment or to a flapwise moment, may be obtained. Multiple positions along the length of the blade 22 may be selected. Furthermore, other data, indicative of loads on the blade 22, may be used to obtain further series of data and, subsequently, further statistical characteristics. Thus, in other examples, data acquired by sub-modules 311-313 may correspond to , e.g. an edgewise force, a flapwise force, a thrust, a deformation, an acceleration, etc.

[0093] Data acquired in the data measurement module 310 may be fed to a statistical processing module 320. The statistical processing module 320 may include corresponding sub-modules 321-323 to perform statistical analysis of the received data. Statistical analysis may result in the calculation of a statistical characteristic for each of the time series of data received. In an example, the statistical characteristic may be a standard deviation, which inventors have found to be particularly sensitive. In other examples, other statistical characteristics, such as an average, median, or variance may be used.

[0094] The statistical processing module 320 may also receive an input from a window duration module 325. The window duration module 325 may be used to define the duration of the measurement period, i.e. the duration of the time series of data. The measurement period may be adjusted depending on, e.g. operating conditions.

[0095] According to an example, the data measuring module 310 may continuously acquire data with a sampling period of, e.g. 40 ms. Such data may then be treated by the statistical processing module 320. The statistical analysis may be carried out for those samples acquired over a measurement period defined by the window duration module 325. In examples, a slidingwindow may be defined. In other words, statistical analysis may be conducted in a continuous manner, e.g. for each sampling period, by considering the samples acquired over the most recent measurement period defined by the window duration module 325.

[0096] Statistical characteristics for the three blades calculated in the statistical processing module 320 may be fed to a comparison module 330. The comparison module 330 may be configured to compare the statistical characteristics to identify potential deviations. Different implementations may be envisaged for the comparison module 330.

[0097] In the example depicted in Figure 5, the comparison module 330 comprises three subtractors, which may be configured to obtain the difference between pairs of statistical characteristics. A first subtractor (uppermost subtractor in Figure 5) may be used to calculate the difference between the statistical characteristic, e.g. standard deviation of the edgewise moment, of the first blade and the statistical characteristic obtained for the third blade. A second subtractor (middle subtractor in Figure 5) may conduct the same operation for the first and the second blade, whereas a third subtractor (lowermost) may take into account the statistical characteristics for the second and the third blade.

[0098] As also shown in Figure 5, the comparison module 330 may also comprise submodules to obtain a percentage value of the difference obtained after each subtractor and a further sub-module to obtain the absolute value of the difference.

[0099] As already mentioned, other criteria may be used to implement the comparison in the comparison module 330. As a first alternative example, the comparison may comprise comparing a statistical characteristic, e.g. the standard deviation, of a first blade, not with each of the statistical characteristics of the other blades, but with an average value of the statistical characteristics of the other two blades. In another example, the statistical characteristic of each blade may be compared with the average of the values of the statistical characteristics of all three blades.

[0100] As also shown in Figure 5, the method may comprise a filter module 340. Such a filter module 340 may be provided to avoid noise so as to increase the reliability of the blade damage detection method. A low pass filter (LPF) may be provided for each of the statistical characteristics.

[0101] Subsequently, a detection module 350 may be provided. In particular, the detection module 350 may comprise comparing the results obtained from the comparison module 330, optionally after being filtered in the filter module 340, with some predetermined thresholds. As an example, a potential defect may be detected if a difference between a statistical characteristic of a blade and a statistical characteristic of another blade reaches a value of,e.g. more than 10% of the value of the corresponding statistical characteristic. As also shown in Figure 5, the process may be repeated continuously and a new set of statistical characteristics may be calculated in case no such defect is identified in the detection module 350.

[0102] In case an abnormal value is detected in the detection module 350, the method may continue to the next module, i.e. a notification and response module 380. In any case, in an example, a further check may be carried out before taking any subsequent action. Thus, as schematically depicted in Figure 5, an operating condition module 360 and a delay module 361 may be provided. The operating condition module 360 may be provided to ensure that the wind turbine 10 is operating normally, i.e. it is not operating under abnormally high wind or facing an event, such as a low voltage ride through (LVRT) event. Such events may lead to abnormal values on the detected loads even in the absence of structural damage in the blade, so the operating condition module 360 may be used to ensure that the wind turbine 10 is operating normally and, accordingly, the abnormal behavior observed can be indeed attributed to a structural defect in a blade. Furthermore, a delay module 361 may be used to ensure such normal operating conditions exists for some time, i.e. to avoid false detections that may occur shortly after exceptional events such as a LVRT event.

[0103] Consequently, as shown in Figure 5, an AND logic gate is provided at the output of the detection module 350, so that a positive result is only obtained if the detected abnormal value in the detection module 350 occurs while the wind turbine 10 operates, and has been operating for a certain predetermined time, substantially under normal operating conditions.

[0104] Regarding the notification and response module 380, different actions may be taken upon detection of a damaged blade 22. In particular, a stopping sub-module 382 may be provided to initiate a controlled stopping of the wind turbine 10 so as to prevent potential catastrophic damage. Furthermore, in the example depicted in Figure 5, an alarm delay block 381 may also be provided to avoid false detections. In particular, the alarm delay block 381 may be configured with a delay such that further action, e.g. stopping of the wind turbine according to the stopping sub-module 382, may not be carried out unless the blade failure detection remains for a certain predetermined minimum time.

[0105] As already described with reference to the flowchart in Figure 4, the different actions implemented in the different modules of Figure 5 may be repeated over a certain observation period of time. In this manner, not only instantaneous differences between the statistical characteristics may be observed, but also trends of such differences may be evaluated.

[0106] Figure 6 schematically illustrates the results 400 of a simulation obtained over a relatively short period of time of about 3 hours. The upper graph 420 illustrates the curves 421- 423 for the values of the standard deviations for the edge moments obtained with a measurement period of 10 minutes for the three blades, i.e. these curves correspond to the results obtained after the statistical processing module 320 in Figure 5. In particular, curve 421 corresponds to blade #1 , curve 422 to blade #2, and curve 423 to blade #3. The graph in the middle shows the results 430 obtained after the comparison module 330 and the filtering module 340. In particular, a first curve 431 is used to represent the difference (in percentage) of the standard deviation for the edge moment in blade #1 and in blade #2. A second curve 432 is used to represent the difference of the standard deviation for the edge moment in blade #1 and in blade #3. Finally, curve 433 represents the difference between the values obtained for blade #2 and blade #3. The bottom graph illustrates the result 440 obtained after the detection module 350 of Figure 5. In this example, a threshold is predetermined such that a blade damage is detected if the difference in any of the previously depicted curves 431-433 exceeds a value of 10%.

[0107] From Figure 6, it is apparent that blades #2 and #3 exhibit a substantially equivalent behavior. Indeed, the standard deviations measured for those blades, as shown in curves 422 and 423, have substantially the same value over the observation period. On the other hand, a clearly different behavior is observed for blade #1 , i.e. in curve 421. Such abnormal behavior of blade #1 can also be detected by looking at the curves of the differences in the middle graph for those curves involving the standard deviation in blade #1 , i.e. curves 431 and 432. Indeed, as shown in the bottom graph, the difference between the standard deviation measured in blade #1 and the other two blades exceeds a value of 10% for most of the considered observation period. Consequently, a structural damage in blade #1 can be derived from the behavior depicted in Figure 6.

[0108] Figure 7 shows another example. In this case, the results obtained when applying a method like the one schematically depicted in Figure 5 over a long period of time. The Figure shown in Figure 7 is equivalent to the middle figure of Figure 6. Hence, Figure 7 also illustrates the differences obtained when subtracting the standard deviations of the edge moment obtained with measurement periods of 10 minutes.

[0109] In Figure 7, curve 510 corresponds to the difference between the standard deviations for blades #2 and #3, whereas curve 520 and curve 530 correspond to the difference between blades #1 and #2, and blades 1# and #3, respectively.

[0110] As shown in Figure 7, all standard deviations are substantially the same, i.e. differences around 0%, up to a certain time, identified as t_1 , of wind turbine operation.Nevertheless, a change is observed at about t_1 , when the start of a progressive, almost linear degradation, can be observed in some of the curves. In particular, curve 510, which concerns blades #2 and #3, remains substantially constant, which is indicative of substantially the same loads being faced by the corresponding blade sensors. On the contrary curves 520 and 530 exhibit a clear downward trend, which is indicative of increasing differences between the loads faced by the blade sensors of blade #1 and the blade sensors in blades #2 and #3. Accordingly, when combining the behavior of the different curves, damage in blade #1 can be deduced.

[0111] Furthermore, the observation of the long term trend in Figure 7 provides increased information. Thus, a significant acceleration of the degradation, i.e. of the structural damage, can be observed after a certain time, t_2, from which degradation becomes substantially exponential. Such a change in the behavior may be indicative of a potentially upcoming catastrophic event. Consequently, potential catastrophic damage may be prevented if stopping the wind turbine 10 as soon as possible, i.e. after time t_1 , once first deviations are observed. On the other hand, after time t_2, the deviation seems to indicate such a significant damage that blade breakage or collapse may occur.

[0112] As shown by the notification and response module 380 of Figure 5, an aspect of the present disclosure may also provide for a method to operate a wind turbine 10, i.e. not only a method for detecting a blade damage. Hence, Figure 8 shows a flowchart of an example of such a method 200 for operating a wind turbine 10 with a rotor 18 having a plurality of blades 22. The method 200 comprises measuring time series of data indicative of loads on a first blade including at least a first time series of data measured with a blade sensor over a measurement period in block 210. Then, in block 220, the method 200 comprises calculating a first statistical characteristic from the measured first time series of data. Subsequently, the method 200 comprises, in block 230, detecting a structural damage of the first blade at least partially based on the calculated first statistical characteristic. Then, block 240 of the method 200 comprises stopping the wind turbine 10 upon detection of the structural damage.

[0113] According to this example, a controlled shutdown of the wind turbine 10 may be implemented based on the observed degradation. By stopping the wind turbine 10, potential catastrophic event, e.g. breakage of the structurally damaged blade 22, may be avoided. Furthermore, a tailored shutdown may be implemented to increase the safety of the maneuver.

[0114] It is to be understood that, while implementing a method for operating a wind turbine according to this aspect of the disclosure, the steps comprising detection of a structural damage in the blade may be carried out according to any of the previously described examples. In other words, the method for operating a wind turbine, which comprises stopping the wind turbine upon detection of a blade damage, may be combined with any of the previouslydescribed methods for detecting such structural damage in a blade. In particular, the method for operation a wind turbine may be combined with methods for detecting a blade damage comprising using sensor data from multiple blades, and / or with methods comprising repeating measurements of series of data and calculating statistical characteristics over certain observation periods to obtain historic or trend behaviors.

[0115] In an example, the method 200 for operating a wind turbine 10 may comprise, prior to stopping the wind turbine 10 upon detection of the structural damage, repeating the measurement of a time series of data indicative of loads on the first blade and the calculation of the first statistical characteristic after a delay period from the previous measurement, and only stopping the wind turbine 10 if structural damage of the first blade is also detected based on the newly calculated first statistical characteristic.

[0116] According to this example, stopping of the wind turbine 10 due to potentially erroneous signals captured at a single cycle may be avoided.

[0117] In still other examples, multiple iterations may be carried out. Hence, in an example, a method for operating a wind turbine may comprise continuously repeating, during an observation period, the measurement of times series of data and the calculation of first statistical characteristics. A trend may be derived from the calculated first statistical characteristics. Then, stopping the wind turbine upon detection of the structural damage may comprise controlling a stopping sequence based on the derived trend.

[0118] Such a method may be implemented over relatively short periods of time, e.g. in the range of hours, and the evolution or trend of the degradation may be analyzed. Such evolution, together with the information on the specific blade sensor showing the degradation, may allow a more informed action. Specifically, this information may be considered to implement a tailored shut down of the wind turbine 10. In other words, stopping of the wind turbine 10 may be carried out according to a sequence that may be adjusted to, e.g. the severity and / or the location, of the detected structural blade damage.

[0119] In an example, the method 200 may comprise stopping the wind turbine by pitching the blades 22 along respective longitudinal axes 34. Furthermore, pitching the blades 22 may comprise pitching the blade 22 with structural damage differently from the other blades.

[0120] According to this example, a gentler treatment may be given to the damaged blade 22 so as to reduce the likelihood of a potential serious event during the stopping maneuver. The specific implementation of the maneuver may depend on the severity and the location of the damage, e.g. on whether the structural damage is more critical for loads in the edgewise or flapwise direction. Thus, in some cases, the damaged blade may be pitched more slowlythan the other blades whereas, in other cases, a faster pitching may be preferred. In still other examples, the damaged blade may not be pitched at all during the stopping sequence.

[0121] Furthermore, in an example, a method for operating a wind turbine 10 may be provided which, after stopping of the wind turbine 10, may provide enhanced mitigation of loads for the damaged blade. In the non-operating state, weight may become the predominant load for the blades. Thus, in order to mitigate loads on the damaged blade, the azimuth angle of the wind turbine rotor 18 may be adjusted such that the damaged blade 22 may be in a substantially vertical position, i.e. pointing upwards or downwards. Once in this position, the wind turbine rotor 18 may be locked to prevent further movements.

[0122] In another aspect of the disclosure, a wind turbine 10 comprising a tower 15, a rotor 18 with one or more blades 22, and a control unit 650 is provided. The control unit 650 is configured for receiving a first time series of data indicative of a load on a first blade 22 over a measurement period and for calculating a first statistical characteristic from the received first time series of data. Furthermore, the control unit 650 is configured for detecting structural damage of the first blade 22 at least partially based on the calculated statistical characteristics.

[0123] In an example, the wind turbine 10 may comprise one or more sensors arranged on the first blade 22 and configured for acquiring the data indicative of a load on the first blade 22 (see sensors 991 , 992 in Figures 3A and 3B). In an example, such sensors may be load sensors and they may be specifically arranged on the first wind turbine blade 22 to implement a method for detecting structural blade damage. In another example, such sensors may be load sensors and they may be arranged on the first wind turbine blade 22 to feed signals for at least another function of the wind turbine 10. Specifically, the load sensors may also be used to provide signals to the blade pitch controller, i.e. the same set of sensors may be used to detect potential structural damage in the blade and to provide pitch angle commands.

[0124] Different types of sensors may be used and a plurality of sensors may be provided. In an example, one or more sensors may be arranged on the first blade 22 and they may be configured for acquiring data indicative of a load on the first blade 22. Specifically, one or more sensors 991 , arranged in the proximity of a leading edge 181 of the first blade 22 may be configured for measuring loads in an edgewise direction.

[0125] Furthermore, in an example, one or more sensors 992, arranged in the proximity of a spar cap 761 , 762 in a downwind or in an upwind portion of the first blade 22, may be configured for measuring loads in a flapwise direction.

[0126] Regarding the position of the sensors 991 , 992 in the spanwise direction of the blade 22, different positions may be selected depending on the specific design of the blade 22. In anexample, the sensors 991 , 992 may be provided in the first half of the blade 22 as measured from the blade root end 171 (see Figure 3A). More particularly, the sensors 991 , 992 may be arranged in the first 10% of the length of the blade 22 as measured form the blade root end 171.

[0127] It is understood that, in wind turbines comprising multiple blades, a suite of sensors may be arranged in each blade or, at least, in more than one blade. In an example, the different suites of sensors may comprise equivalent sensors, i.e. the same number of sensors, the sensors being configured to detect the same loads. This may also include redundant sensor systems.

[0128] Accordingly, in wind turbines comprising, e.g. three blades, it is understood that the control unit 650 may be configured to replicate the above-mentioned features for all the blades or, at least, for some of the blades apart from the first blade.

[0129] Figure 9 schematically illustrates an example of a layout of components in a wind turbine 10 with a control unit 650 configured for detecting blade damage according to an example. In particular, the wind turbine 10 may comprise three blades 221-223, and each blade may be equipped with three sensors 991-993. As an example, each blade may comprise a sensor 991 arranged in the proximity of the leading edge 181 for detecting edgewise loads and two sensors 992, 993 arranged in the proximity of the spar caps for detecting flapwise loads. In an example, the sensors 991-993 may comprise fiber Bragg grating (FBG) sensors, so optical fiber communication lines 620 may be provided between the respective sensors 991- 993 and break out boxes 601-603 arranged for each blade. A further optical communication line 630 may be provided between the break out boxes 601-603 and an optoelectronic converter 640, which may be used to convert the optical signals into electrical signals for further processing. Subsequently, electrical signals may be feed to a control unit 650. In the control unit 650, a method such as the one described with reference to Figure 5 may be implemented.

[0130] As understood by the skilled person, Figure 9 is provided only for illustrative purposes and different variations, including with different type and number of sensors, or different type of communication lines may be envisaged in other examples of the disclosure.

[0131] The control unit 650 may be embodied as an independent controller, i.e. a controller intended only to conduct the detection of a structural defect in any of the blades 22. Alternatively, the control unit 650 may be embodied in another controller such as a pitch system controller 80 or the wind turbine controller 36.

[0132] In an example, the control unit 650 may further be configured for receiving data from at least one of the sensors arranged on the first blade and for generating commands to controla pitch angle of the first blade during normal operation. According to this example, the same sensors 991 , 992 and control unit 650 may be employed for at least two different purposes. Thus, they may be used to detect a structural blade damage as described in the present disclosure and they may also be used to assist during the blade pitch control during normal operation of the wind turbine 10. Furthermore, such sensors 991 , 992 and control unit 650 may also be configured for other functions such as for ice detection. Accordingly, data measured by the sensors 991 , 992 may be fed to the control unit 650, which may use such data in multiple ways to accomplish several control objectives. By using the same sensors 991 , 992 and control unit 650, a more cost effective wind turbine 10 may be obtained.

[0133] This written description uses examples to disclose the teaching, and also to enable any person skilled in the art to practice the teaching, including making and using any devices or systems and performing any incorporated methods. The patentable scope is defined by the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements with insubstantial differences from the literal languages of the claims. Aspects from the various examples described, as well as other known equivalents for each such aspects, can be mixed and matched by one of ordinary skill in the art to construct additional examples and techniques in accordance with principles of this application. If reference signs related to drawings are placed in parentheses in a claim, they are solely for attempting to increase the intelligibility of the claim, and shall not be construed as limiting the scope of the claim

Claims

29CLAIMS1 . A method (100) for detecting a structural damage of a first blade (22) of a wind turbine (10), the wind turbine (10) comprising a rotor (18) with one or more blades (22), the method (100) comprising: measuring time series of data indicative of loads on the first blade (22) including at least a first time series of data measured with a blade sensor (991 , 992) over a measurement period; calculating a first statistical characteristic from the measured first time series of data; and detecting the structural damage of the first blade (22) at least partially based on the first calculated statistical characteristic.

2. The method (100) of claim 1 , wherein the first statistical characteristic is a standard deviation of the measured data over the measurement period.

3. The method of any of claims 1 or 2, comprising calculating a second statistical characteristic from the measured first time series of data, the second statistical characteristic being different from the first statistical characteristic, and further comprising detecting the structural damage of the first blade based on the first statistical characteristic and on the second statistical characteristic.

4. The method (100) of any previous claims, wherein measuring the time series of data comprises measuring while the wind turbine (10) is in operation.

5. The method (100) of any previous claim, wherein measuring data indicative of loads on the first blade (22) comprises measuring data indicative of loads acting in an edgewise direction and / or in a flapwise direction of the first blade (22), specifically wherein calculating the first statistical characteristic comprises calculating a statistical characteristic of an edgewise bending moment and / or of a flapwise bending moment.

6. The method (100) of any previous claim, comprising continuously repeating the measurement of time series of data and the calculation of the corresponding first statistical30 characteristic during an observation period, wherein detecting the structural damage of the first blade (22) at least partially based on the first calculated statistical characteristic comprises analyzing an evolution of a value of the first statistical characteristic over the observation period and / or analyzing a rate of change of the value of the first statistical characteristic over the observation period.

7. The method (100) of claim 6, wherein the structural damage of the first blade (22) is detected when a variation of the value of the first statistical characteristic within the observation period exceeds a first threshold, and / or when the rate of change of the value of the first statistical characteristic exceeds a second threshold within the observation period.

8. The method (100) of any previous claim, comprising: measuring time series of data indicative of loads on a second blade (22) including at least a first time series measured with a blade sensor over the measurement period; calculating a second statistical characteristic from the measured first time series of data; wherein detecting the structural damage of the first blade (22) comprises comparing the first statistical characteristic with the second statistical characteristic.

9. The method (100) of claim 8, wherein the first statistical characteristic and the second statistical characteristic correspond to a same statistical property.

10. The method (100) of any of claims 8 or 9, comprising continuously repeating, during an observation period, the measurement of time series of data for the first and the second blade, the calculation of the corresponding first and second statistical characteristics, and the comparison of the calculated first statistical characteristic with the second statistical characteristic, wherein detecting the structural damage of the first blade (22) comprises analyzing an evolution of the comparison over the observation period.11 . The method (100) of any previous claim, comprising: measuring time series of data in all the blades (22) over the measurement period, the data being indicative of loads on the corresponding blade (22); calculating corresponding statistical characteristics from the measured time series of data;wherein detecting the structural damage of the first blade (22) is based on the statistical characteristics calculated for all the blades (22), specifically based on a comparison of the statistical characteristic calculated for the first blade (22) with the statistical characteristics calculated for the other blades (22).

12. The method (100) of any previous claim, wherein detecting the structural damage of the first blade (22) at least partially based on the first statistical characteristic comprises determining if a difference of a value of the first statistical characteristic with respect to a predetermined value for the first statistical characteristic exceeds a predetermined threshold.

13. A method (200) of operating a wind turbine (10), the wind turbine (10) comprising a rotor (18) with a plurality of blades (22), the method comprising: detecting a structural damage of a first blade (22) with a method according to any of the previous claims, and stopping the wind turbine (10) upon detection of the structural damage.

14. The method (200) of claim 13, wherein stopping the wind turbine (10) comprises pitching the blades (22) along respective longitudinal axes (34), and further wherein pitching the blades (22) comprises pitching the first blade (22) with structural damage differently from the other blades.

15. A wind turbine (10) comprising a tower (15), a rotor (18) with one or more blades (22), and a control unit (650), the control unit (650) being configured for: receiving a first time series of data indicative of a load on a first blade (22) over a measurement period; calculating a first statistical characteristic from the received first time series of data; detecting structural damage of the first blade (22) at least partially based on the first calculated statistical characteristic.

Images