System for monitoring the vibrations of a wind turbine

Abstract

The invention proposes a system for monitoring the vibrations, in particular the natural frequency, of a tower (1) of a wind turbine (2), comprising vibration sensors (3) associated with the tower (1) and comprising an evaluation device (4) for processing vibration sensor data. In order to be able to make reliable statements about damage to the tower (1), it is proposed that a plurality of vibration sensors (3) are arranged annularly around the tower axis (6) in three or more tower cross-sectional planes (5) along the tower axis (6), wherein the evaluation device (4) stores, in addition to the natural frequency (fi), vibration amplitudes (A) associated with the respective tower cross-sectional planes (5) at observation times (Ti) and identifies structural damage in the tower (1) by way of comparing the natural frequencies (fi) and the amplitude ratios (A) over an observation period.

Description

[0001] for monitoring the vibrations of a wind turbine

[0002] Technical field

[0003] The invention relates to a system for monitoring the vibrations, in particular the natural frequency, of a tower of a wind turbine, comprising vibration sensors assigned to the tower and an evaluation device for processing vibration sensor data, wherein several vibration sensors are arranged in a ring-shaped or spiral pattern around the tower axis in three or more tower cross-sectional planes, and wherein the evaluation device stores vibration amplitudes assigned to the respective tower cross-sectional planes for observation periods in addition to the natural frequency.

[0004] State of the art

[0005] A wind turbine converts the kinetic energy of the wind into electrical energy and feeds it into a power grid. The tower of the wind turbine is the supporting component of the entire turbine. Wind forces, rotor blade forces, and the weight of all components located at the top of the tower are transferred through it to a foundation. When the wind turbine is in operation, wind forces and the rotation of the rotor cause the tower to vibrate. Excitation at the tower's natural frequency must be avoided to prevent damage from resonant vibrations, which can ultimately lead to the tower's destruction. Therefore, monitoring the tower's natural frequency is known from CN 113279920 A. This involves a data acquisition unit and a control unit, the data acquisition unit being designed to detect vibration signals from the tower and transmitting these signals to the control unit.The control unit monitors the received vibration signal and compares it with a preset threshold value in order to shut down the wind turbine if the threshold value is exceeded.

[0006] CN 116772924 discloses a modal system for testing wind turbines. This system comprises a group of stress sensors mounted on the tower walls to detect the vibration amplitude and the loads applied to the tower. The system aims to determine the tower's current vibration frequency more precisely based on the acquired data, thereby enabling long-term monitoring. The measurement of the tower's vibration amplitude is described to analyze its response to different wind speeds and directions. The vibration amplitude and the loads acting on the tower are related to the wind field parameters to perform modal analyses of the natural frequencies.

[0007] Currently, no reliable information is available regarding the actual lifespan of wind turbines. Estimating their lifespan requires relying on empirical data and average values. With current methods, it is not possible to pinpoint potential damage, thus preventing targeted maintenance measures. This poses the risk of a sudden total failure of the wind turbine and can also lead to the premature decommissioning of a still fully functional tower, thereby increasing operating costs.

[0008] Description of the invention

[0009] The invention is therefore based on the objective of providing a system of the type described above, which enables the detection and, if necessary, the localization of defects occurring in the tower using simple means and also allows conclusions to be drawn about the probability of tower failure. The invention solves this problem by having the evaluation device detect structural damage in the tower by comparing the natural frequencies and amplitude ratios over an observation period. By monitoring the vibrations and comparing the amplitudes, it is possible, according to the invention, to detect emerging or existing damage to the tower.

[0010] The invention comprises several vibration sensors arranged in a ring or spiral pattern along the tower axis in three or more tower cross-sectional planes. An evaluation unit calculates and records the natural frequency and the vibration amplitudes of the tower cross-sectional planes over observation periods and compares the natural frequencies and amplitude ratios over an observation period. If current measurement results deviate from previously stored measurement results beyond a permissible tolerance, this indicates detected structural damage in the tower. If this deviation is only present in one of the upper tower cross-sectional planes, this indicates that the damage is located in the area below the tower cross-sectional plane where the deviation first occurs and above the tower cross-sectional plane that does not yet show any deviation.This allows the damage to be localized, which reduces manual diagnosis and repair time. This also helps to shorten the time required for routine inspections and only fully entrust them to the system once it is mature.

[0011] The tower should be equipped with at least one temperature sensor. The temperature sensor data, along with the vibration amplitudes, should be stored for each observation period and incorporated into a comparison of amplitude ratios over that period. This allows for the consideration of temperature-dependent changes in the tower's vibration behavior, which, mechanically, resembles a beam fixed at one end. Similarly, the tower should be equipped with at least one wind sensor. The wind sensor data, along with the vibration amplitudes, should be stored for each observation period and incorporated into a comparison of amplitude ratios over that period. All sensor data can be transmitted remotely, either via cable or a wireless network, to a central control unit.

[0012] The vibration sensors can be installed inside the towers, particularly on the inner tower wall. Since the natural frequency and amplitude ratio are used as reference values, the system is suitable for all known tower construction materials (e.g., steel, concrete, wood, plastic, etc.). The vibration sensors are preferably, but not exclusively, implemented as piezoelectric crystals, as these, once installed, resonate with the component and output a vibration frequency and amplitude ratio, which are then evaluated by comparison with older measurements stored in the database. A warning can be issued for any deviations within narrow limits. Above a threshold with a larger deviation, the wind turbine is automatically shut down. The temperature sensors can be implemented as conventional temperature sensors.The vibration is excited by the wind forces and any imbalance in the rotors.

[0013] The system according to the invention monitors any existing damage and, if applicable, its progression by means of sensors arranged in a ring or spiral along the tower axis. By arranging the vibration sensors in a ring shape in at least three tower cross-sectional planes spaced apart along the tower axis—with four vibration sensors per tower cross-sectional plane, each offset from the others by 90°—it is possible to significantly improve the measurement results with regard to the ratios of vibration amplitudes and natural frequency. Improved directional information for the vibrations is also obtained. A change in the ratios over the observation period indicates the onset or progression of damage to the tower.A comparison of vibration amplitudes at the same wind speed but different wind directions indicates damage, particularly in onshore wind turbines, if deviations are found. For offshore turbines, the flow-dependent wave motion must also be considered. Furthermore, this arrangement allows for the testing of various vibration scenarios, which can lead to inaccurate conclusions when considering only the natural frequency. In addition, if a maximum vibration amplitude is exceeded, even with an otherwise intact tower, and the exceedances are therefore caused by external factors such as waves, failure of one or more rotor blades, etc., the turbine can be shut down to prevent tower damage.

[0014] The natural frequency and its harmonics, as well as their amplitudes, are determined from the measured values ​​using the Fourier transform. If the scatter of the measured values ​​is too large from the outset, either a sensor is defective (and can be replaced), or there may be an inhomogeneity in the component, indicating a potential source of damage. The frequency closest to the theoretical natural frequency is used. During this process, the temperature is always taken into account, as it influences both frequency and amplitude. After this adjustment, the natural frequency is determined and stored as a value in a database. The amplitude of each sensor at this natural frequency, taking into account wind speed and temperature, is also stored in a database. The main focus is on comparing the amplitudes under the respective operating conditions and observing how they change.

[0015] After initial installation, the natural frequency of the wind turbine, and in the case of offshore turbines, also that of the mounting structure (monopiles, jackets, floating elements, etc.), is calculated from measurement results and, where possible, compared with the natural frequency and amplitudes of an identical component. For onshore turbines, impending foundation failure can also be detected. This allows any existing damage to be identified or estimated. The system according to the invention can also be installed during operation. However, since the natural frequencies or amplitude ratios of individual wind turbines (including or excluding mounting structures) can differ due to tolerances, the values ​​determined during initial installation are used for further analysis.

[0016] The ratio of the vibration amplitudes is preferably determined as a function of wind speed and wind direction. Information regarding wind speed and direction can be provided by suitable sensors assigned to the wind turbine.

[0017] Measurements are preferably performed continuously. The measured data are stored in a database and compared with older data, enabling automated monitoring as well as on-site and remote monitoring. Therefore, as soon as the natural frequency or the amplitude ratio changes significantly at a specific temperature, it can be inferred that damage, such as the beginning of cracking or signs of disintegration, is present. This damage is reported to the operator, and if necessary, the wind turbine is shut down to prevent further damage or total structural failure.

[0018] In most cases, the damaged area can be localized and appropriate repair measures can be carried out in a targeted manner.

[0019] With the system according to the invention, however, it is not necessary to make inaccurate remaining service life estimates, since each tower can be monitored individually, and any damage occurring can be detected by a change in natural frequency and / or changes in amplitude ratios. A dynamic response can be taken to these changes on a case-by-case basis, thus avoiding the need to replace an entire wind farm and therefore increasing the overall service life of the wind farm.

[0020] Preferably, the evaluation unit can monitor the vibration amplitudes assigned to the respective tower cross-sectional planes over observation periods, since their change (especially increasing monotonic change) indicates progressive damage. Particular attention is paid to the amplitudes of the initially determined natural frequency, as these change more significantly than the frequencies themselves. In particular, a drift in the frequencies is detected when the amplitudes change noticeably.

[0021] Brief description of the invention

[0022] The invention is illustrated in the drawing as an example. It shows

[0023] Fig. 1 shows a wind turbine with a system according to the invention in view, Fig. 2 shows the wind turbine from Fig. 1 in section along line 11-11, Fig. 3 shows, above, a diagram of an example oscillation at a

[0024] Tower cross-sectional plane and below it a Fourier transform of the signal and

[0025] Fig. 4 shows the deflection with an intact 4a tower and with a tower showing signs of failure 4b.

[0026] Ways to implement the invention

[0027] The system for monitoring the vibrations, in particular the natural frequency, of a tower 1 of a wind turbine 2 comprises vibration sensors 3 assigned to the tower 1 and an evaluation unit 4 for processing vibration sensor data. Several vibration sensors 3 are arranged in a ring or spiral pattern around the tower axis 6 in three or more tower cross-sectional planes 5. In addition to the natural frequency, the evaluation unit 4 stores vibration amplitudes A assigned to the respective tower cross-sectional planes 5 for observation periods Ti (a suitably short measurement period) and detects structural damage in the tower by comparing the natural frequencies fi and the amplitude ratios over an observation period.

[0028] Tower 1 can be assigned at least one temperature sensor 7, whereby the temperature sensor measurement data, together with the vibration amplitudes, are stored for observation periods and included in the comparison of the amplitude ratios over the observation period. Likewise, tower 1 can be assigned at least one wind sensor, whereby the wind sensor measurement data, together with the vibration amplitudes, are stored for observation periods and included in the comparison of the amplitude ratios over the observation period. In the present case, the wind sensor data are provided, for example, by the control system 8 of the wind turbine 2, which already includes a corresponding wind sensor.

[0029] The vibration sensors 3 are preferably arranged inside the towers, in particular on the inner tower wall (Fig. 2). In the illustrated embodiment, four vibration sensors 3 are provided for each tower cross-sectional plane, each offset from the others by 90°.

[0030] The observation periods can be selected periodically and / or load-dependently. Preferably, the evaluation unit 4 calculates the vibration amplitudes assigned to the respective tower cross-sectional planes 5 for observation periods, stores these, and detects and locates structural damage in the tower by comparing the amplitudes determined over an observation period.

[0031] Various types of instruments are used to measure vibrations in towers.

[0032] Three suitable vibration sensors. Firstly, there would be

[0033] Accelerometers, such as piezoelectric accelerometers, which use piezoelectric materials to convert mechanical movements into electrical signals, are robust, highly sensitive, and suitable for a wide frequency bandwidth. Capacitive accelerometers measure changes in capacitance caused by movements of the internal structure. They offer high accuracy and are well-suited for low frequencies. MEMS (microelectromechanical systems) are small, cost-effective sensors used in many applications. They offer good sensitivity and are available in various frequency ranges. Seismometers, velocity sensors, laser vibrometers, and / or inclinometers can also be used to measure vibrations in towers.The selected sensors must be able to detect the expected vibration frequencies of the building, be sensitive enough to detect the vibrations to be measured, be durable enough to reliably perform long-term measurements, and should require minimal installation and calibration effort.

[0034] The natural frequency fi of a tower 1 can be determined from a vibration measurement, as explained for example in Fig. 3, as follows: The evaluation unit 4 stores vibration amplitudes A assigned to the respective tower cross-sectional planes 5 for observation time intervals Ti. For this purpose, vibrations are first recorded over a suitably short measurement period (in the example around 10 s), which should by no means be considered limiting.

[0035] Figure 3 shows a representation of the Fourier transform of a simulated oscillation. The upper diagram depicts the time-domain signal, which shows the simulated oscillation with noise. The lower diagram shows the frequency-domain signal, i.e., the result of the Fourier transform. The frequencies with the highest amplitudes correspond to the natural frequencies of the system. In this example, distinct peaks at 1 Hz and 3 Hz are visible, corresponding to the simulated oscillation frequencies. For the measurement setup and data acquisition, suitable vibration sensors 3 are installed at appropriate locations in the tower, particularly in the upper section, as the vibration amplitude is greatest there. Subsequently, vibration data is acquired and recorded over a specific period. The measurement includes both natural wind-generated and artificial rotor-generated vibrations.Ensure that the vibration sensors 3 are correctly calibrated. Measurements should be taken over a sufficient period to capture all relevant frequencies.

[0036] Environmental conditions during the measurement, such as wind, temperature and, if applicable, humidity, are documented along with the vibration data.

[0037] The raw data is filtered to remove noise and unwanted frequencies. Typically, a high-pass filter is used to eliminate low-frequency noise and a low-pass filter to eliminate high-frequency noise. The data is also checked for anomalies and corrected if necessary. Erroneous measurements are logged but discarded for further processing.

[0038] The preprocessed vibration data are subjected to a Fourier transform (FFT), which transforms the time-domain data into the frequency spectrum of the vibrations. The peaks are then identified within this frequency spectrum (Fig. 3 below).

[0039] The dominant frequencies are then determined and compared with the resonant frequencies of similar towers. The natural frequency is typically the lowest dominant frequency. Finally, the results are compared with theoretical models and / or previous measurements to validate the measurement. To determine the tower's natural modes and frequencies in more detail, a modal analysis of the FFT transform can also be performed. Figures 4a and 4b schematically show the deflection of an intact tower (4a) and a tower with incipient failure (4b). The defect leads to a change in the deflection curve, possibly a kink in the curve. The illustration is intended to indicate a small deflection below the defect and a significantly larger one above it, thus allowing the defect to be spatially localized. The angular position of the defect can be determined via the respective sensor.

Claims

Patent claims 1. System for monitoring the vibrations, in particular the natural frequency, of a tower (1) of a wind turbine (2), comprising vibration sensors (3) associated with the tower (1) and an evaluation unit (4) for processing vibration sensor data, wherein several vibration sensors (3) are arranged in a ring or spiral shape around the tower axis (6) in three or more tower cross-sectional planes (5), wherein the evaluation unit (4) stores vibration amplitudes (A) associated with the respective tower cross-sectional planes (5) for observation periods (T), characterized in that the evaluation unit (4) detects structural damage in the tower (1) by comparing the natural frequencies (ft) and the amplitude ratios (A) over an observation period.

2. System according to claim 1, characterized in that at least one temperature sensor (7) is assigned to the tower (1), wherein the temperature sensor measurement data together with the vibration amplitudes are stored for observation periods and are included in the comparison of the amplitude ratios over the observation period.

3. System according to claim 1 or 2, characterized in that at least one wind sensor is assigned to the tower (1), wherein the wind sensor measurement data together with the vibration amplitudes are stored for observation periods and are included in the comparison of the amplitude ratios over the observation period.

4. System according to one of claims 1 to 3, characterized in that sensor data is transmitted to a central unit via remote transmission, in particular via cable or a radio network, or is processed directly in the evaluation unit (4) in the wind turbine (2).

5. System according to one of claims 1 to 4, characterized in that the vibration sensors (3) are arranged inside the tower (1), in particular on the inner wall of the tower.

6. System according to one of claims 1 to 5, characterized in that at least four are arranged one behind the other around the tower cross-sectional plane (5). Vibration sensors (3) are provided arranged at 90° intervals along the tower axis (6).

7. System according to one of claims 1 to 6, characterized in that the observation times (Ti) are selected periodically and / or load-dependently.

8. System according to one of claims 1 to 7, characterized in that the evaluation device (4) stores vibration amplitudes (A) assigned to the respective tower cross-sectional planes (5) along the tower axis (6) for observation periods and, by comparing the vibration frequencies and especially their amplitudes determined over an observation period, detects and locates structural damage in the tower (1) and issues a warning in case of minor deviation and, if necessary, automatically shuts down the wind turbine in case of major deviation.

Images