Determining life of wind turbine

Abstract

The invention describes a method of determining a wear contribution (Ncyc) of at least one bearing of a wind turbine (41), the method comprising: performing a computational simulation of a mechanical behavior of at least part of the wind turbine (41) comprising the bearing for each of a plurality of given operating conditions (28, 29) of the wind turbine across a predetermined simulation time window (8); determining, on the basis of the simulation, at least one mechanical torque (Mx, My), in particular two mechanical torques, which act on the bearing, across the simulation time window (8); determining a time sequence (22) of at least one force-dependent variable (21) across the simulation time window (8) on the basis of the mechanical moments (Mx, My) across the simulation time window (8); determining a time sequence (24a,...) of stress-dependent variables (23) across the simulation time window (8) on the basis of the force-dependent variable time sequence (22) and in particular on the basis of the pitch position (14); the stress-dependent variable time series (24a,...) is evaluated in order to obtain a wear contribution (Ncyc) dependent on the respective operating conditions (28, 29) and the simulation duration.

Description

Technical Field

[0001] The invention relates to a method of determining a contribution of damage / deterioration / wear of at least one bearing of a wind turbine and a corresponding arrangement. Further, the invention relates to a method of estimating a wear indicator of a bearing of a wind turbine and a corresponding arrangement. Further, the invention relates to a method of controlling a wind turbine comprising at least one bearing and a corresponding arrangement. Still further, the invention relates to a wind turbine. Background Art

[0002] US2014288855 A1 discloses a control strategy for occasionally increasing the rated power of a wind turbine during its design life, the control strategy comprising: measuring operating parameters of components of the wind turbine when the wind turbine operates at a design rated speed and a design rated torque; and calculating actual accumulated damage to the components based on the measured values ​​of the operating parameters.

[0003] US2022228559 A1 discloses a method for controlling a wind turbine connected to a power grid, the method comprising receiving a state estimate of the wind turbine via a controller. The method further comprises: determining, via the controller, a current condition of the wind turbine using at least the state estimate, the current condition defining a set of condition parameters of the wind turbine.

[0004] US2021123416 A1 discloses a method for operating a wind turbine. The wind turbine is operated during an operation period according to a control strategy. During the operation period, a sensor signal is received from a sensor measuring an operating parameter of the turbine. Based on the sensor signal, a model is used to obtain a modeled fatigue value.

[0005] A wind turbine comprises a plurality of bearings which rotatably support, for example, a rotating shaft or a rotor blade. The blade bearings are designed to rotatably support the rotor blades and to allow the blade pitch angle to be adjusted as required.

[0006] During operation of the wind turbine, one or more bearings may experience wear and / or degradation and / or at least local damage. The bearings are subject to mechanical stresses due to forces generated by impinging wind or due to movement of components. Blade pitch bearings may be designed and operated under the assumption that all bearings will experience the conditions presented in the design simulation. Conventionally, blade bearings may be designed and risk models may be calculated based on the design simulation.

[0007] In reality, the bearings may experience less severe conditions or may experience more severe conditions many times or in many cases than those in the design simulation. The result of this difference may be that at the end of the design turbine life, the blade bearings may still have residual capacity; or in another case, the rotor blade bearings have an increased risk of failure before the end of the design turbine life.

[0008] Traditionally, predicted failure rates might only be compared with retrospective actual failure rates and based on this, forecasts of warranty costs / expenditures are continually updated. However, the knowledge gained from such programs might only influence the structural design or design life estimates of future wind turbines.

[0009] Traditionally, simulations might be calibrated / validated on prototype wind turbines, but would not take into account the individual wind conditions and operating hours of each wind turbine (sometimes customers take turbines out of service, during which period the bearings will experience significantly less wear).

[0010] However, it has been observed that the simulated remaining life of the bearing does not reflect the true remaining life of the bearing in all situations or circumstances.

[0011] Therefore, a method and a corresponding arrangement for determining a wear contribution of at least one bearing of a wind turbine may be needed, a method for estimating a wear indicator of a bearing of a wind turbine may be needed, and a method and a corresponding arrangement for controlling a wind turbine taking into account the wear indicator of the bearing may be needed. Summary of the invention

[0012] Embodiments of the invention may enable the contribution to wear to be determined taking into account individually experienced external conditions, such as wind conditions, in particular taking into account wake versus free wind, average wind on a given side, etc. Embodiments of the invention may enable the remaining life of a bearing to be determined with greater reliability than conventionally known methods.

[0013] According to an embodiment of the present invention, a method for determining a wear (damage / degradation / ) contribution of at least one bearing of a wind turbine is provided, the method comprising: performing a computational simulation of the mechanical behavior of at least a part of the wind turbine including the bearing for each of a plurality of given operating conditions of the wind turbine across a predetermined simulation time window; determining, based on the simulation, at least one mechanical moment (e.g., load), in particular two mechanical moments for a blade pitch bearing, acting on the bearing, across the simulation time window; determining, based on the mechanical moments across the simulation time window, at least one force-related parameter (e.g., rolling element force, ball force) time series across the simulation time window; determining, based on the force-related parameter time series (and in particular based on the pitch position), a stress-related parameter (e.g., contact pressure) time series across the simulation time window; and evaluating the stress-related parameter (e.g., contact pressure) time series in order to obtain a wear contribution (e.g., Ncyc) associated with the respective operating condition and simulation duration.

[0014] The method can be applied to one or more rotor blade pitch bearings. For each rotor blade pitch bearing, in particular two moments can be determined.

[0015] The method may be implemented by software and / or hardware and may be executed or controlled by a corresponding arrangement for determining a wear contribution. The wear contribution may indicate how much wear and / or damage and / or degradation a bearing has experienced when operating the wind turbine under respective operating conditions. The wear may also indicate stress, and the wear contribution may indicate the contribution of stress experienced by the wind turbine bearing. For example, the wear contribution may be indicated as the number of standard stress cycles.

[0016] The bearing may be designed to withstand a specific amount of cumulative wear or a specific amount of cumulative (standard or equivalent) stress cycles. The stress cycles may relate to, for example, the distribution of contact pressures experienced by the bearing, or in particular the distribution of contact pressures experienced by the raceways of the bearing. The stress cycles may be caused by rolling elements within the bearing contacting the raceways of the bearing. The wear contribution may be assigned different physical units, or simply as a numerical value, for example, the number of standard stress cycles experienced. The given stress cycle (e.g., indicating the contact pressure across time) may be converted into a specific number of standard or equivalent stress cycles, as known in the art. The standard stress cycle may be a cycle having a maximum value of contact pressure of a specific value (such as 3000MPa).

[0017] The bearing may comprise a rotor blade pitch adjustment system bearing, or for example a bearing of a main shaft or secondary shaft of a wind turbine.The pitch bearing may allow rotation substantially about the longitudinal axis of the rotor blade in order to adjust the blade pitch angle.

[0018] The computational simulation may utilize, for example, a simulation software package.

[0019] The behavior and loads of the turbine can be simulated, for example, in BHawC. For the bearing calculations, Ansys, Julia and / or Matlab can be used, but other software may also be employed.

[0020] By taking into account eg the geometrical / material configuration of the wind turbine blade and the wind turbine blade bearings, the mechanical behavior may be simulated.

[0021] The plurality of given operating conditions may include external operating conditions and / or internal operating conditions of the wind turbine, as will be explained below.

[0022] The predetermined simulation time window may be set according to the specific application, and may be set, for example, in the range of 10 minutes to 15 minutes. The simulation time window may be sampled or subdivided into a plurality of simulation time steps, the simulation time step having a duration, for example, between 0.01 seconds and 0.1 seconds.

[0023] A normal operating time series may be simulated, for example, at 25 Hz, which means that the duration of each time step is 1 second / 25 = 0.04 seconds. In some special cases, the frequency may be, for example, up to 100 Hz, resulting in a time step duration of 0.01 seconds.

[0024] The mechanical moments may be calculated from the forces generated by the impinging wind and the geometrical configuration of the rotor blade and the blade bearing. In particular, two mechanical moments may be determined, for example, one in the X direction and one in the Y direction, wherein the X direction and the Y direction may bridge a plane corresponding to or parallel to a plane bridged or defined by the bearing rings. The bearing may comprise an inner raceway, a rolling element in contact with the inner raceway, and an outer raceway, the rolling element also in contact with the outer raceway. For example, the rolling element may comprise one or more rolling balls or rolling cylinders. The inner raceway may be arranged at the inner ring and likewise the outer raceway at the outer ring. The mechanical moment may act on one of the raceways relative to the other. The rolling element is arranged between the raceways and in contact with the raceways.

[0025] For example, the force-related variables may correspond to forces experienced or exerted by rolling element, e.g. ball forces acting between the respective rolling element and the inner and / or outer raceway. Forces acting on the inner raceway as well as forces acting on the outer raceway may be determined and taken into account. In particular, the forces transmitted or exerted or experienced by all rolling elements between the two raceways should be considered separately with respect to the effect of the forces on the inner and outer raceways. Primarily, the forces may be considered to affect the (structural) integrity of the raceways. The rolling elements may be considered not to have been degraded by the operation of the wind turbine.

[0026] The force-related variable time series may correspond to a time series (multiple time steps) related to (a variable of) a force acting on the inner raceway and / or the outer raceway. The respective forces may be derived based on the determined mechanical moments and the geometrical configuration of the rotor blade and / or the respective bearing. The computational simulation may require input data related to a definition of the geometrical configuration of the part of the wind turbine being simulated, the wind turbine comprising the bearings and in particular also comprising a rotor blade or a plurality of rotor blades connectable or coupled to a main shaft of the wind turbine. The force-related variable time series may correspond to values ​​of the force-related variable over a plurality of time steps, the plurality of time steps forming a simulation time window.

[0027] The force-related time series may in particular be considered in conjunction with the corresponding pitch position (time series) in order to determine a stress-related parameter time series. The stress-related parameters may, for example, indicate mechanical stresses or strains, in particular contact pressures and / or subsurface stresses or strains, to which the raceways and / or rolling elements are subjected across a plurality of time steps within a simulation time window. During the execution of the simulation, the movement of the rolling elements within the bearing and / or the movement of the inner raceways and / or outer raceways may be modeled. Due to the uncertainties associated with the initial positioning of the rolling elements, multiple simulations may be performed for a given operating condition, wherein different initial positions of the rolling elements are employed.

[0028] The stress-related parameter time series may, for example, include one or more stress cycles having the same or different maximum stress values, in particular contact pressure values ​​and / or subsurface stresses or strains. The plurality of stress cycles (in particular stress cycles with different heights) may be combined and converted into, for example, a (single) parameter or number representing a wear contribution.

[0029] The final result may represent or indicate failure rates and / or risk distributions. The model may create a stress cycle "Markov matrix" that may be used to calculate risk distributions. The Markov matrix is ​​a simple matrix containing the number of cycles or occurrences for a given combination of mean stress and stress range.

[0030] In particular, the time series of the stress-related parameters (in particular, contact pressure) can be converted or evaluated to obtain a number of standard or equivalent stress cycles. This number can reflect the wear or damage or degradation that the bearing is simulated to have experienced after completing operation with the operating conditions as defined. The equivalent stress cycles can be used for visualization and rapid evaluation.

[0031] When the wear contributions are determined in the manner described above, the wear contributions can advantageously be utilized in order to estimate the respective accumulated wear that the bearings of the wind turbine have been subjected to after the wind turbine has been operated over a certain operating interval. Subsequently by utilizing a combination, in particular a summation of the wear contributions, for example the remaining life of the wind turbine bearings can be determined.

[0032] According to an embodiment of the present invention, the force-related parameter indicates a mechanical force acting at at least one position of the raceway of the bearing, in particular a rolling element force or a ball force, wherein the stress-related parameter indicates a stress at at least one position of the raceway of the bearing, in particular a contact pressure; and / or the method further comprises: evaluating the stress-related parameter time series together with the pitch angle position of the rotor blade supported by the bearing, so as to adopt a cycle counting (in particular a rainflow counting) method.

[0033] The rolling element forces acting on the inner raceway and the outer raceway of the bearing can be determined. During operation during the simulation time interval, the locations at which the rolling element forces act may change. The simulation can take into account the respective locations at which the forces act on the raceways. In this sense, the simulation can determine the position-dependent forces on the raceways and based on the position-dependent forces it can also be determined that the respective stress-related parameters are also position-specific. Thereby, the simulation results can be more reliable.

[0034] The Rainflow Counting algorithm can be used to calculate the fatigue life of a bearing in order to convert a load sequence of bearing stresses (e.g., a contact pressure time series) into an equivalent set of constant amplitude stress reversals, in particular resulting in a specific number of standard stress cycles. The method can sequentially extract smaller interruption cycles from the sequence, which models the material memory effect observed using stress-strain hysteresis cycles. The Rainflow Counting algorithm was developed by Tatsuo Endo and N. Matsuishi in 1968. According to an embodiment of the present invention, the Rainflow Counting algorithm can be used.

[0035] According to an embodiment of the present invention, evaluating the time series of the stress-related parameter (for example, contact pressure) includes: assigning different values ​​of the stress-related parameter in the time series to different intervals; counting the number of occurrences in each interval; considering relevant intervals, combining the number of occurrences, in particular summing the number of occurrences weighted by the interval number, so as to obtain the wear contribution.

[0036] The stress-related parameter across the simulation time window can in particular be evaluated to obtain a single value in order to simplify the result database. The interval division allows to classify the different values ​​of the stress-related parameter, thereby simplifying the method, in particular reducing the storage requirements. By summing the number of occurrences in each interval weighted by the interval number (for example, indicating the level of the respective value of the stress-related parameter), the normalization of the wear contribution can be obtained. The overview procedure can assume that the wear or damage or degradation of a first number of stress cycles with a first intensity or height is essentially equivalent to the wear or damage or degradation of a second number of stress cycles with a second height or intensity. It is thereby easier to accumulate the contributions of wear or damage or degradation caused by different intensities or amounts of stress (in particular contact pressure).

[0037] According to an embodiment of the invention, the bearing is at least part of a bearing system of a blade pitch angle adjustment system for adjusting the pitch angle of a rotor blade; and / or wherein the part of the wind turbine comprises at least one of: at least one rotor blade; a bearing system of a blade pitch angle adjustment system; at least one raceway of at least one bearing; at least one rolling element or rolling ball of at least one bearing; and / or wherein the simulation time window is between 5 minutes and 15 minutes, in particular, approximately 10 minutes.

[0038] According to other embodiments, more components of the wind turbine may be considered during the simulation. Other embodiments may allow modeling of other components of the wind turbine, in particular other bearings. The rolling element elements, in particular rolling element balls, may be arranged between an outer raceway and an inner raceway of the bearing, the outer raceway may be formed by an inner surface of an outer ring, and the inner raceway may be formed by an outer surface of an inner ring.

[0039] According to an embodiment of the invention, the operating condition contains a value indicating at least one of: at least one external or environmental parameter to which the considered wind turbine is subjected, in particular comprising at least one of: wind speed, in particular a minimum and / or maximum value and / or an average value and / or a standard deviation across the simulation time window; turbulence, in particular a minimum and / or maximum value and / or an average value and / or a standard deviation across the simulation time window.

[0040] Both external and environmental parameters can influence the forces acting on different parts of the simulated element. When also taking into account the minimum and / or maximum and / or mean and / or standard deviation across different parameters, a respective distribution of the simulation results can also be obtained, which distribution can be advantageously exploited for later use of the simulation results for assessing the operating period of an actual wind turbine. Both wind speed and / or turbulence can significantly influence the forces acting on different components. In other embodiments, such as for offshore wind turbines, wave parameters can also be taken into account, such as, for example, wave height, wave frequency or wave wavelength.

[0041] According to an embodiment of the invention, the operating condition comprises a value indicating at least one of: at least one wind turbine-internal operating parameter relating to the wind turbine under consideration, in particular comprising at least one of: a control setting; a control scheme; a rotational speed; a pitch angle of the rotor blades including the bearing; a rotor blade load, in particular a flap-directed moment and / or an edge-directed moment; a wind turbine power output.

[0042] The operating parameters internal to the wind turbine may include all operating parameters of the wind turbine, which parameters are different from external operating parameters, such as environmental parameters. The control settings may, for example, include a definition of a rated power output, a rated rotational speed, a rated wind speed, a rotational speed-power curve, etc. The considered control settings may also affect the control of the wind turbine during the simulation time window, thereby also affecting the generated forces. The control scheme may, for example, include a definition of normal operation, a definition of limited operation, etc. The control settings and / or control scheme may also indicate to what extent the control of the wind turbine varies under full load operation versus partial load operation (e.g., below rated wind speed). The control settings and / or control scheme may involve the definition of one or more reference curves (such as a power-rotational speed curve, a wind speed-pitch angle curve, etc.).

[0043] Thereby, typical operating conditions of a wind turbine can be captured and taken into account for the simulation.

[0044] According to an embodiment of the invention, across the simulation time window the pitch angle and / or the rotational speed is kept constant or varies due to simulated control of the wind turbine according to one of a plurality of wind turbine control schemes.

[0045] Although certain internal operating parameters (such as, in particular, pitch angle and / or rotational speed) may vary during the simulation of a given operating condition, at least external operating conditions related to wind speed and / or wind turbulence may remain constant. Thereby, the simulation may reflect a more realistic behavior of the wind turbine in dependence on at least the external environmental conditions.

[0046] According to an embodiment of the invention, the wear contribution indicates a degree of damage and / or a degree of degradation and / or a degree of wear and / or a remaining life and / or a consumed life, wherein the wear contribution is particularly expressed as a number of standard (or equivalent) stress cycles (Ncyc) of the bearing, the standard stress cycle being particularly dependent on at least wind speed and wind turbulence.

[0047] The number of standard stress cycles may, for example, correspond to the number of stress cycles with a height of 3000 MPa experienced at at least one location on at least one raceway of the bearing, thereby simplifying the analysis by comparing the simulation output with conventionally used parameters, for example, for expressing component life.

[0048] According to an embodiment of the invention, the simulation is run multiple times for each respective operating condition for different initial values ​​of the position of one or more rolling elements or rolling balls in order to obtain a distribution of wear contributions associated with the respective operating conditions and the simulation duration, wherein in particular statistical parameters of the distribution of the wear contributions are determined, in particular the mean value and / or minimum value and / or maximum value and / or standard deviation.

[0049] Taking into account different initial values ​​of the rolling element may improve the accuracy of the resulting estimated wear contribution.Furthermore, the statistical quantities (eg statistical moments) may advantageously be used for also providing an idea about the distribution of the resulting wear contribution.

[0050] According to an embodiment of the invention, the method comprises storing the wear contribution (Ncyc) associated with the respective operating condition (e.g. simulation duration and / or construction details about the wind turbine) in an electronic storage device, in particular as a simulation result that can be requested based on an input of the operating condition of interest.

[0051] The electronic storage device may be, for example, a storage device of a wind turbine controller or a storage device of a wind farm controller or a storage device of some other separate entity, which storage device is accessible to all wind turbines of the wind farm or at least to the wind farm controller. The simulation results may represent data in a simplified manner, which data may be accessed for estimating the wear or in particular the number of standard stress cycles to which at least one bearing has been subjected after a certain operating period of an actual wind turbine.

[0052] According to an embodiment of the present invention, a method for estimating a wear index (Ncyc) of a bearing of a wind turbine is provided, the method comprising: recording operating conditions during operation of the wind turbine spanning a plurality of operating duration time intervals; for the respective operating conditions, retrieving or interpolating a damage contribution for each operating duration time interval from stored simulation results, in particular a damage contribution obtained according to one of the aforementioned embodiments; and summing the damage contributions in order to obtain a wear-related index.

[0053] The method may be performed, for example, by a portion of a wind turbine, such as a wind turbine controller or a wind turbine evaluation module. In other embodiments, the method may be performed by a wind farm monitoring system or a wind farm controller. The operating conditions may be recorded, for example, in a wind turbine electronic storage device or a wind farm monitoring storage device.

[0054] When the wind turbines are operated during at least one operation duration time interval at respective actual operating conditions (the actual operating conditions being different from any operating conditions included in the stored simulation results), the respective wear contribution may be interpolated from adjacent wear contributions corresponding to the operating conditions included in the simulation results. Thereby, the wear indicator may be estimated in a very fast manner.

[0055] The wear indicator may in particular indicate the number of standard cycles to which the bearing or at least one location on at least one raceway of the bearing has been subjected.

[0056] The wear index (e.g. Ncyc) can be calculated at many independent points (evenly) distributed across all raceways, which is used to calculate the risk of failure of the bearing at each point. By calculating the risk of failure as a combination of "system risk", the risk of failure of the complete bearing can be determined. The risk of failure can then be converted to theoretical life or wear based on an accepted failure rate.

[0057] According to an embodiment of the present invention, a method for controlling a wind turbine is provided, the wind turbine comprising at least one bearing, in particular a pitch adjustment system bearing, the method comprising: estimating a wear indicator of the bearing according to the aforementioned embodiment; adjusting the control of the wind turbine so as to increase or decrease bearing usage according to the estimated wear-related indicator.

[0058] When the wear indicator of the bearing is estimated in an accurate manner, the control of the wind turbine can be adjusted, in particular in order to increase the power output and / or reduce the degradation of loads or wear of the wind turbine, or in order to achieve a combination of such objectives.

[0059] According to an embodiment of the present invention, the method further comprises comparing the estimated wear indicator with a design wear indicator; adjusting the control comprises: if the estimated wear indicator indicates less wear than the design wear indicator, adjusting the control of the wind turbine to increase bearing usage, in particular including increasing power output; and / or, if the estimated wear indicator indicates more wear than the design wear indicator, adjusting the control of the wind turbine to reduce bearing usage, in particular including reducing power output.

[0060] The estimated wear index may be given, for example, as an estimated number of standard cycles and the design wear index may be given as a design number of standard cycles to which the bearing has been subjected.

[0061] Increasing bearing usage may involve changing the pitch angle at a higher rate and / or applying changes in the pitch angle more frequently than normal. Reducing bearing usage may involve changing the pitch angle at a lower rate and / or applying changes in the pitch angle less frequently than normal.

[0062] It should be understood that the features disclosed, described, illustrated or provided for the method for determining the wear contribution of at least one bearing of a wind turbine, alone or in any combination, may also be applicable, alone or in any combination, to the arrangement for determining the wear contribution of at least one bearing of a wind turbine according to embodiments of the present invention, and vice versa.

[0063] According to an embodiment of the invention, an arrangement for determining a wear contribution of at least one bearing of a wind turbine is provided, the arrangement comprising: a processor adapted to perform a computational simulation of the mechanical behavior of at least a part of the wind turbine including the bearing for each of a plurality of given operating conditions of the wind turbine across a predetermined simulation time window; the processor adapted to determine, based on the simulation, across the simulation time window at least one mechanical moment, in particular two mechanical moments, acting on the bearing; the processor adapted to determine, based on the mechanical moments across the simulation time window, at least one force-related variable (e.g. a ball force) time series across the simulation time window; the processor adapted to determine, based on the force-related variable time series (and in particular based on the pitch position), a stress-related variable (e.g. a contact pressure) time series across the simulation time window; the processor adapted to evaluate the stress-related variable (e.g. a contact pressure) time series in order to obtain a wear contribution (e.g. Ncyc) associated with the respective operating condition and the simulation duration; the arrangement further comprising: an electronic storage device for storing the wear contribution (e.g. Ncyc) associated with the respective operating condition.

[0064] The arrangement may be implemented, for example, by a computing system, which may not be part of the wind turbine. The simulation may be performed offline and may require other inputs besides the operating conditions, such as mechanical / geometric / material / design configuration of the modelled components.

[0065] According to an embodiment of the present invention, a wind turbine is provided, which comprises: at least one bearing, in particular a pitch adjustment system bearing; a monitoring device, which is used to monitor operating conditions during operation of the wind turbine across multiple operating duration time intervals (in particular, a recording device for recording the operating conditions); a retrieval device, which is used to retrieve or interpolate the damage contribution of each operating duration time interval from stored simulation results for the respective operating conditions; a processor, which is suitable for summing the damage contributions to obtain a wear index; and a controller, which is suitable for adjusting the control of the wind turbine according to the estimated wear index to increase or decrease the bearing usage.

[0066] The aspects defined above and further aspects of the invention are apparent from the examples of embodiment to be described hereinafter and are explained with reference to the examples of embodiment. The invention will be described in more detail hereinafter with reference to examples of embodiment but to which the invention is not limited. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Embodiments of the present invention will now be described with reference to the accompanying drawings. The present invention is not limited to the illustrated or described embodiments.

[0068] Figure 1 Data considered for determining a wear contribution of at least one bearing of a wind turbine according to an embodiment of the invention are schematically illustrated.

[0069] Figure 2 and Figure 3 Example results of wear contribution obtained according to embodiments of the present invention are illustrated.

[0070] Figure 4 Examples of wear contributions for different operating conditions obtained according to embodiments of the invention are schematically illustrated.

[0071] Figure 5 An example of monitored operational data such as that obtained according to an embodiment of the present invention for use in evaluating a wind turbine is illustrated; and

[0072] Figure 6 A method of controlling a wind turbine according to an embodiment of the invention is schematically illustrated, further comprising a wind turbine according to an embodiment of the invention and an arrangement according to an embodiment of the invention. DETAILED DESCRIPTION

[0073] exist Figure 1 The method 10 schematically illustrated in FIG. 1 determines a wear contribution of at least one bearing of a wind turbine. Figure 1 A set of load curves 1, an example of determined rolling element forces 2, an example of a time series 3 of force-related parameters, an example of a pitch position 4, an example of a stress cycle determined according to an embodiment of the present invention (marked with reference numeral 5), an example of a time series 6 of stress cycles and an example of performing rain flow counting 7 according to an embodiment of the present invention are illustrated.

[0074] Thereby, a computational simulation of the mechanical behavior of at least part of the wind turbine including the bearing is performed for each of a plurality of given operating conditions of the wind turbine spanning a predetermined time window 8, the time window 8 being between Figure 1 Partial View Figure 1 In the illustrated embodiment, the simulation time window 8 comprises 300 simulation time steps. The simulation time window 8 may, for example, range between 5 minutes and 15 minutes, approximately 10 minutes.

[0075] Based on the simulation, at least one mechanical moment Mx1, My1, ..., across the simulation time window 8, is determined, the mechanical moment Mx1, My1 acting on the first bearing of the first rotor blade. Further, respective bending moments Mx2, My2, Mx3, My3 of two other rotor blades are determined. The respective moments Mx, My represent moments in the X direction and in the Y direction acting on the respective bearing of the respective rotor blade, wherein the XY plane is in particular located in a plane defined by the outer ring and the inner ring of the respective bearing or is parallel to the plane.

[0076] The method 10 involves creating a database of a plurality of blade bearing raceway fatigue calculation results, also referred to hereinafter as respective wear contributions of the bearings of the wind turbine. Each calculation comprises a simulation of the wind turbine at a given wind turbulence and an average wind speed. Thus, the wind turbulence together with the wind speed constitutes an example of an external operating condition to which the wind turbine is subjected. The definition of the operating condition may comprise further external and / or internal operating parameters.

[0077] In the illustrated embodiment, the simulation results in a 10-minute load (shown here as Figure 1 The time series of the moment in part 1 of the drawing and the pitch position in part 4 of the drawing. The respective x-axis or abscissa 11 thus indicates the time step. The ordinate 12 of part 1 of the drawing indicates the moment and the ordinate 13 of part 4 of the drawing indicates the pitch position. The pitch position is given as a pitch position distribution 14. Figure 1As can be seen in Figure part 4, the pitch position 14 varies over the simulation time window 8. This may be partly due to wind turbulence involving wind with varying wind speeds. Depending on the varying wind speed due to the turbulence, the pitch angle of the wind turbine is also controlled, which is governed by the adjusted control settings or control scheme set during the simulation time window 8.

[0078] For each time step, the loads on the bearing (e.g., expressed as moments) are evaluated and the contribution to fatigue damage is determined. This can be achieved by fitting a six-dimensional response plane, where the six moments (Mx and My for each of the three blades) are converted to ball force distributions, as Figure 1 2 of the accompanying drawings. Thus, the axes 15, 16 represent the respective moments Mx, My, while the ordinate 17 indicates the rolling element forces, in particular the ball forces, caused according to a given combination of said moments. Thus, a ball force function 18 is created or determined which depends on the different moments acting on the bearing.

[0079] The ball force function 18 is determined for each rolling element position (defined as circumferential angular position along axis 19 in part 3 of the drawing) and is dependent on time (as indicated by axis 20 in part 3 of the drawing). The ordinate 21 in part 3 of the drawing again illustrates the ball force. Figure 1 In part 3 of the drawing, a time series of force-related variables, in particular ball forces, is determined, which is marked with reference numeral 22. The time series 22 comprises the respective ball forces at a plurality of points in time, which are dependent on the circumferential position of the rolling element balls of the bearing.

[0080] The ball force time series 22 is considered in combination with the pitch position 14 in order to derive stress-related variables (in particular, Figure 1 6 of the accompanying drawings). Therein, the abscissa 11 again indicates the time step, and the ordinate 23 indicates the contact pressure. As can be seen from the accompanying drawings, the time series of the stress-related parameter includes a plurality of stress cycles 24a, 24b, 24c, 24d, 24e, 24f, which have different values ​​or heights. One stress cycle 24c is explained in more detail in the accompanying drawings, part 5. The stress cycle 24c starts at a specific time step with a value of substantially 0, then rises to a maximum value of about 1800 MPa and then decreases again to substantially 0. Each such stress cycle can be converted into a specific number of standard stress cycles, for example, into a specific number of cycles with an amplitude of 3000 MPa.

[0081] As Figure 1As a final step of the method schematically illustrated in , a rainflow counting procedure can be applied, as schematically illustrated in part 7 of the accompanying drawings, where the axis marked with σ indicates stress and the axis marked with t indicates time. Thus, a normalization of the indicators of wear can be achieved.

[0082] In short, according to Figure 1 The load input is combined with the response plane to obtain a ball force time series. Further, the pitch position is combined with the ball force time series to obtain a stress time series. Rainflow counts can be derived from the stress time series.

[0083] like Figure 1 The results of the method schematically illustrated in can cover all possible operating wind conditions of a wind turbine in a grid of a given resolution. Each calculation can use as input a set of 10-minute time series of high-frequency data, which are converted into times representing the cumulative fatigue damage for a given time series. The grid points contain multiple simulations with the same wind conditions, but the results may be different due to stochasticity, turbulence and other effects. Figure 2 An example of this is illustrated in . The X axis 26 thus indicates a seed for determining an initial position (eg the position of a rolling element ball) or a seed for determining an instantaneous wind speed for a given turbulence attributable to operating conditions of wind speed = 10 m / s and turbulence = 0.1. Figure 2 It can be seen in FIG. 1 that the randomness of certain data or the uncertainty of certain statistics of data (such as turbulence) leads to a plurality of resulting values ​​for the number of standard cycles (as an example of wear contribution) shown with crosses 28 a , 28 b , . . . .

[0084] The number of results at each grid point must be significant to ensure that it correctly represents the statistical distribution of the accumulated fatigue damage under the given conditions.

[0085] Figure 3 Illustrated are exemplary simulation results (distributions), ordinate 27 indicating the number of standard cycles (indicating the contribution of wear), axis 28 indicating wind speed, axis 29 indicating wind speed turbulence, both being examples of defined operating conditions.

[0086] The data cloud 30 thus indicates the wear contribution of all combinations of wind speed and wind turbulence, wherein for each combination of wind speed and turbulence, in particular, a distribution of the respective wear contribution can be assigned, in particular a distribution of the number of standard cycles can be assigned. Figure 3 The data described in may be stored in an electronic storage device, for example as simulation results that may be accessed via input of operating conditions of interest.

[0087] To perform the required fatigue calculations may be computationally intensive, and the amount of work or calculation time may depend on the resolution of the mesh and the number of calculations for the mesh points. There may be a trade-off between accuracy of the results and calculation time.

[0088] When a sufficient number of calculations have been performed, the resulting combined database can be converted into an n-dimensional response plane, an example of which is shown in Figure 4 . Here, axis 28 again indicates wind speed, axis 29 again indicates wind turbulence, and ordinate 27 indicates the number of standard cycles (as an example of wear contribution). Here, an estimated number 30 of standard cycles is given in the form of a surface diagram, in particular indicating the average number of standard cycles.

[0089] The simulation may further include the results of the minimum / maximum / standard deviation of the number of standard cycles, for example, the results may be obtained from Figure 3 The data illustrated in FIG. 3 are derived in FIG.

[0090] Figure 5 Explains the application of Figure 4 , for evaluating the operation of a wind turbine. Thereby, axis 28 again indicates the wind speed experienced by the wind turbine and axis 29 again indicates the turbulence actually experienced by the wind turbine. For each operating condition defined by wind speed and turbulence, the relevant wear contribution can be derived by looking up the respective values ​​from the simulation results 30 or by interpolating the respective values ​​from adjacent values ​​of slightly different operating conditions.

[0091] Thus, applying the simulation results 30 to the considered wind turbine provides an estimated wear indicator 31 that the wind turbine is predicted to have experienced or has been determined to have experienced. Thus, the simulation results 30 are used to quickly interpolate the average cumulative damage that the considered wind turbine has experienced for given conditions.

[0092] By assigning the result of each of the previous calculations (number of equivalent fatigue cycles) to the relevant combination of input values ​​(in this case wind speed and turbulence), a response plane is created, which then forms the basis for an interpolation table. From the table, the output for any combination of wind speed and turbulence can be quickly interpolated, such as Figure 5 By densely partitioning the input parameter space and using the table to interpolate the number of equivalent cycles for each, a table such as Figure 4 Surface 30 as illustrated in FIG.

[0093] For a given turbine, the table or simulation results 30 may be adapted for given input combinations such that certain combinations never occur while other combinations occur multiple times. Figure 5The operating conditions in are exemplary conditions experienced during three years of operation of an example wind turbine. Figure 5 In the example illustrated in , the average number of cycles (average Ncyc) is used, but the 95% upper percentile or any statistically safe value desired may be used instead. The method is applicable to statistical data (maximum, minimum and average values ​​over 10 minutes) and can be very fast (once a simulation results database is created).

[0094] The wind turbines may record statistical data about operating conditions, which may be accessed via the wind farm communication network. Embodiments of the present invention may apply the determination of wear contribution or cumulative wear to all turbines in the wind farm. The actual utilization of the bearing experienced by each individual bearing specific to the wind turbine may then be determined.

[0095] After a significant operating time (preferably at least 80% of a full operating year), a clear picture or diagnosis of the given bearing can be obtained. Further, it is possible to predict the remaining life (the unused life of the bearing at the time of removal if the turbine continues to operate with the same controller settings). This information can be used for future turbine designs and / or for adjusting the model of the given turbine in order to release (or increase) warranty reserves.

[0096] Among other things, the determined information may be used for each individual wind period to manually adjust / optimize the controller inputs to enable utilization of potential excess capacity in the bearings for the purpose of increasing annual energy production, reducing loads on key components, or other benefits.

[0097] therefore, Figure 6 An exemplary cycle for optimization according to an embodiment of the present invention is illustrated. The scheme 40 illustrates a wind turbine 41, which may also be configured according to an embodiment of the present invention. In a method block or feature block 42, operating data of the wind turbine 41 is captured. Thus, the wind turbine continuously generates data, which is captured and stored in a database. Each turbine experiences unique wind conditions and wave conditions based on location, spacing, positioning, etc. within a given site. When commissioning based on the given conditions experienced by the specific and independent wind turbine, optimal controller settings are achieved.

[0098] In a next method step 43, the data for a given operating time (eg from the complete duration of the operating life of the turbine) are evaluated (in particular using / accessing data such as Figure 4 30) in order to determine the remaining life of each relevant component by using the above-mentioned response plane method.

[0099] The remaining life of the bearing is calculated in a schematic method step 44, whereby the simulation results 30 (see also Figure 4 ). Thus, said step 44 involves performing a method for estimating a wear indicator of a bearing of a wind turbine, wherein said wear indicator is in particular indicative of the number of equivalent fatigue cycles experienced by at least one bearing.

[0100] In an evaluation step 45, it is evaluated whether the remaining life (obtained by performing the method for estimating the wear index) is greater or less than the expected life or the designed life. Thus, by comparing the remaining life of the component with the expected life of the turbine or of the component, it can be determined whether the local conditions of the turbine allow for more aggressive or less aggressive operation.

[0101] If the remaining lifetime is greater than the expected lifetime, a switch is made to branch 46 which leads to an adjustment step 47. Thereby, the controller is optimized for increased bearing usage (e.g. involving increased AEP, load reduction, reduction of loads on components with high risk / failure rate (e.g. main rotor bearings), etc.). Thereby, the wind turbine may be manually or automatically commissioned with the aim of increasing, for example, annual energy production or reducing utilization of other critical components. The overall risk of the turbine may be reduced and warranty reserves thereby released.

[0102] If the estimated remaining life is less than the expected life, branch 48 is directed to an adjustment step 49. In the adjustment step 49, the controller is optimized for reduced bearing usage (e.g., involving reducing AEP, accepting higher loads on components (e.g., main rotor bearings), thereby reducing pitch stroke, etc.). Thus, the wind turbine may be adjusted manually or automatically with the goal of protecting critical components by reducing AEP or higher loads on other components. Alternatively, it may be decided that increased risk may be permitted for a particular turbine.

[0103] New data is collected after commissioning the controller, and after a given period of time when a sufficient amount of data has accumulated, a new optimization of the controller can be performed. These processes can be repeated.

[0104] The methodology is particularly applicable to blade bearings of a wind turbine, but may also be used for main bearings, yaw bearings, gearboxes and / or similar components of a wind turbine.

[0105] Embodiments of the present invention may provide one or more of the following benefits or advantages:

[0106] The methodology may enable processing of lifetime data from a large number of turbomachines (fleet assessment).

[0107] This can be used to determine the remaining life of the bearings on each individual wind turbine and thereby enable optimization of the controller and / or site / fleet either manually or through embedded software at the individual wind turbine level.

[0108] All conceivable operating conditions may have been considered in advance, and for these said operating conditions, the respective wear contributions may be available. Based on this, the experienced component (bearing) loads may be quickly calculated, for example based on 10 minutes of statistical data. Thus, embodiments of the present invention may directly convert operating data into fatigue damage.

[0109] Furthermore, an overview of the damage of each individual bearing in the fleet can be quickly provided, since all conceivable conditions have been simulated in advance and how they affect the components (eg bearings) calculated.

[0110] If the simulation diagnoses that a reserve of capacity of the respective bearing is used, the wind turbine may be controlled to increase the AEP and / or reduce the failure rate.

[0111] Figure 1 The method explained in the above may be implemented by an arrangement 50 according to an embodiment of the present invention (see Figure 6 ) is executed, the arrangement 50 is used to determine the wear contribution of at least one bearing of a wind turbine 41, the arrangement comprising: a processor, the processor being adapted to perform a computational simulation of the mechanical behavior of at least a part of the wind turbine 41 including the bearing, the computational simulation being performed for each of a plurality of given operating conditions of the wind turbine across a predetermined simulation time window 8; the processor being adapted to determine at least one mechanical moment Mx, My, in particular two mechanical moments, acting on the bearing based on the simulation across the simulation time window 8; the processor being adapted to determine at least one mechanical moment Mx, My, in particular two mechanical moments, acting on the bearing based on the simulation across the simulation time window The mechanical torque is used to determine at least one force-related parameter 21 time series 22 across the simulation time window 8; the processor is suitable for determining a stress-related parameter 23 time series (24a, ...) across the simulation time window 8 based on the force-related parameter time series and in particular based on the pitch position; the processor is suitable for evaluating the stress-related parameter 23 time series (24a, ...) in order to obtain a wear contribution (Ncyc) associated with the respective operating conditions and the simulation duration; the arrangement also includes an electronic storage device, which is used to store the wear contribution (Ncyc) associated with the respective operating conditions.

[0112] A wind turbine 41 according to an embodiment of the present invention (see Figure 6) comprises: at least one bearing, in particular a pitch control system bearing (for a rotor blade 41a or 41b or 41c); a monitoring device 42, the monitoring device being used to monitor the operating conditions during the operation of the wind turbine spanning a plurality of operating duration time intervals; a retrieval device, the retrieval device being used to retrieve or interpolate the damage contribution for each operating duration time interval from the stored simulation results for the respective operating conditions; a processor, the processor being suitable for summing the damage contributions in order to obtain a wear index; a controller, the controller being suitable for adjusting the control of the wind turbine to increase or decrease the bearing usage according to the estimated wear index.

[0113] It should be noted that the term "comprising" does not exclude other elements or steps, and "a" or "an" does not exclude a plurality. In addition, elements described in association with different embodiments may be combined. It should also be noted that the reference numerals in the claims should not be interpreted as limiting the scope of the claims.

Claims

1. A method for determining a wear contribution (Ncyc) of at least one bearing of a wind turbine (41), the method comprising: performing a computational simulation of the mechanical behavior of at least a portion of the wind turbine (41) including the bearing for each of a plurality of given operating conditions (28, 29) of the wind turbine over a predetermined simulation time window (8); Based on the simulation, determining at least one mechanical moment (Mx, My), in particular two mechanical moments, acting on the bearing, across the simulation time window (8); Determining a time series (22) of at least one force-related variable (21) across the simulation time window (8) based on the mechanical torque (Mx, My) across the simulation time window (8); Determining a time series (24a, ...) of stress-related variables (23) spanning the simulation time window (8) based on the time series (22) of the force-related variables and in particular based on the pitch position (14); The stress-related variable time series (24a, . . .) are evaluated in order to obtain a wear contribution (Ncyc) which is dependent on the respective operating conditions (28, 29) and the simulation duration.

2. The method according to the preceding claim, in, The force-related variable (22) is indicative of a mechanical force, in particular a rolling element force or a ball force, acting on at least one location of a raceway of the bearing, wherein the stress-related variable (23) indicates a stress, in particular a contact pressure, at at least one location of a raceway of the bearing; and / or The method further comprises: The stress-related variable time series (24a, . . . ) is evaluated together with the pitch angle position (14) of the rotor blade supported by the bearing, using a cycle counting, in particular rainflow counting (7) method.

3. A method according to any one of the preceding claims, wherein: Evaluating the time series (24a, ...) of the stress-related parameters (23) comprises: assigning different values ​​of the stress-related parameter in the time series to different intervals; Count the number of occurrences in each interval; The numbers of occurrences are combined taking into account the relevant intervals, in particular the numbers of occurrences weighted by the interval number are summed in order to obtain the wear contribution.

4. The method according to any one of the preceding claims, in, The bearing is at least part of a bearing system of a blade pitch angle adjustment system for adjusting the pitch angle of a rotor blade (41a); and / or Wherein, the part of the wind turbine (41) comprises at least one of the following: at least one rotor blade (41a); Bearing system for blade pitch angle adjustment system; at least one raceway of at least one bearing; at least one rolling element or rolling ball of at least one bearing; and / or The simulation time window is between 5 minutes and 15 minutes, and in particular is approximately 10 minutes.

5. A method according to any one of the preceding claims, wherein: The operating condition contains a value indicating at least one of the following: The at least one external parameter or environmental parameter to which the wind turbine is subjected that is taken into account comprises in particular at least one of the following: wind speed (28), in particular the minimum and / or maximum and / or average and / or standard deviation across the simulation time window; Wind turbulence (29), in particular the minimum and / or maximum and / or mean and / or standard deviation across said simulation time window.

6. A method according to any one of the preceding claims, wherein: The operating condition contains a value indicating at least one of the following: At least one wind turbine internal operating parameter related to the wind turbine under consideration, in particular comprising at least one of the following: Control settings; Control scheme; Rotation speed; Rotor blade loads, in particular flap-directed moments and / or edge-directed moments; Wind turbine power output; A pitch angle of the rotor blade including the bearing.

7. A method according to any one of the preceding claims, wherein: Across the simulation time window (8), the pitch angle (14) and / or the rotational speed is kept constant or varies due to simulated control of the wind turbine according to one of a plurality of wind turbine control schemes.

8. The method according to any one of the preceding claims, in, The wear contribution (Ncyc) indicates the degree of damage and / or the degree of degradation and / or the degree of wear and / or the remaining life and / or the consumed life, Therein, the wear contribution is in particular expressed as the number of normal or equivalent stress cycles (Ncyc) of the bearing and is in particular dependent on at least the wind speed and the wind turbulence.

9. A method according to any one of the preceding claims, wherein: The simulation is run a plurality of times for each of the respective operating conditions (28, 29) for different initial values ​​of the position of one or more rolling element or rolling element balls in order to obtain a distribution (30) of the wear contribution associated with the respective operating condition and the simulation duration, In this case, in particular statistical variables of the distribution of the wear contribution are determined, in particular a mean value and / or a minimum value and / or a maximum value and / or a standard deviation.

10. A method according to any one of the preceding claims, comprising: The wear contribution (Ncyc) associated with the respective operating condition is stored in an electronic memory device, in particular as a simulation result which can be requested by inputting the operating condition of interest.

11. A method for estimating a wear index (Ncyc) of a bearing of a wind turbine (41), the method comprising: recording (42) operating conditions during operation of the wind turbine spanning a plurality of operating duration time intervals; retrieving or interpolating the impairment contribution for each operating duration time interval from simulation results (30) stored for said respective operating conditions obtained according to the preceding claim; The damage contributions are summed to obtain the wear-related indicator.

12. A method of controlling a wind turbine, the wind turbine comprising at least one bearing, in particular a pitch adjustment system bearing, the method comprising: Estimating a wear indicator of said bearing according to the preceding claim; Control of the wind turbine is adjusted to increase or decrease bearing usage based on the estimated wear-related indicator.

13. The method according to the preceding claim, further comprising: comparing the estimated wear index with the design wear index; Adjusting the control includes: If the estimated wear indicator indicates less wear than the design wear indicator, adjusting control of the wind turbine (41) for increased bearing usage, in particular including increasing power output; and / or If the estimated wear indicator indicates more wear than the design wear indicator, control of the wind turbine (41) is adjusted for reduced bearing usage, particularly including reducing power output.

14. An arrangement for determining a wear contribution of at least one bearing of a wind turbine (41), the arrangement comprising: A processor, the processor being adapted to: performing a computational simulation of the mechanical behavior of at least a portion of the wind turbine (41) including the bearing for each of a plurality of given operating conditions of the wind turbine over a predetermined simulation time window (8); Based on the simulation, determining at least one mechanical moment (Mx, My), in particular two mechanical moments, acting on the bearing, across the simulation time window (8); determining a time series (22) of at least one force-related variable (21) across the simulation time window (8) based on the mechanical torque across the simulation time window; Determining a time series (24a, ...) of stress-related variables (23) spanning the simulation time window (8) based on the time series of the force-related variables and in particular based on the pitch position; evaluating the time series (24a, ...) of the stress-related variables (23) in order to obtain the wear contribution (Ncyc) associated with the respective operating conditions and simulation duration; An electronic storage device is provided for storing a wear contribution (Ncyc) associated with the respective operating condition.

15. A wind turbine (41), comprising: at least one bearing, in particular a pitch adjustment system bearing; a monitoring device for monitoring an operating condition during operation of the wind turbine spanning a plurality of operating duration time intervals; a retrieval device for retrieving or interpolating the impairment contribution for each operating duration time interval from the stored simulation results by accessing the electronic storage means of the arrangement according to claim 14 for the respective operating condition; a processor adapted to sum the damage contributions to obtain the wear indicator; A controller is adapted to adjust control of the wind turbine to increase or decrease bearing usage based on the estimated wear indicator.

Images