WIND ENERGY SYSTEM AND METHOD FOR IDENTIFYING LOW-FREQUENCY OSCILLATIONS IN AN ELECTRICAL SUPPLY NETWORK

Abstract

Provided is a method for identifying low-frequency oscillations, in particular subsynchronous resonances, in an electrical supply network, wherein the electrical supply network has a line voltage with a nominal line frequency, comprising the steps: detecting at least one electrical signal of the electrical supply network as at least one test signal and filtering and/or transforming the at least one detected test signal into at least one check signal, temporal derivation of the at least one check signal or difference formation of temporally separated values of the check signal, in order to obtain a gradient signal in each case, identifying the presence of a low-frequency oscillation if the gradient signal or at least one of the gradient signals meets a predetermined check criterion, in particular at least one predetermined check limit is exceeded.

Claims

1. A method for identifying low-frequency oscillations in an electrical supply network having a line voltage associated with a nominal line frequency, the method comprising: detecting at least one electrical signal of the electrical supply network as at least one test signal; filtering and/or transforming the at least one test signal into at least one check signal; obtaining a gradient signal by: temporally deriving the at least one check signal, or obtaining a difference of temporally separated values of the at least one check signal; determining whether the gradient signal meets a predetermined criterion; and in response to determining that the gradient signal meets the predetermined criterion, identifying a presence of a low-frequency oscillation.

2. The method as claimed in claim 1, wherein: the predetermined criterion is met in response to the gradient signal exceeding a gradient maximum value, and determining whether the gradient signal meets the predetermined criterion includes: determining whether the gradient signal exceeds the gradient maximum value at least once, or determining whether the gradient signal exceeds the gradient maximum value at least for a predetermined minimum period.

3. The method as claimed in claim 1, wherein obtaining the difference of temporally separated values of the at least one check signal includes obtaining a difference between a maximum and minimum value of the at least one check signal in a check period.

4. The method as claimed in claim 1, wherein: the at least one test signal are line voltages of the electrical supply network in three phases, the at least one test signal are current fed into the electrical supply network in three phases, or the at least one test signal are the line voltages of the electrical supply network in three phases and the current fed into the electrical supply network in three phases.

5. The method as claimed in claim 4, wherein: transforming the at least one test signal into the at least one check signal includes transforming the line voltages of the electrical supply network in three phases and the current fed into the electrical supply network in three phases into a voltage signal, an active power signal and a reactive power signal as a voltage check signal, active power check signal or reactive power check signal, temporally deriving the at least one check signal includes temporally deriving the voltage check signal, the active power check signal and the reactive power check signal to obtain a voltage gradient signal, an active power gradient signal and a reactive power gradient signal, respectively, or obtaining the difference of the temporally separated values of the at least one check signal includes obtaining a difference of temporally separated values of the voltage check signal, the active power check signal and the reactive power check signal to obtain the voltage gradient signal, the active power gradient signal and the reactive power gradient signal, respectively, the voltage gradient signal, the active power gradient signal and the reactive power gradient signal are checked for the presence of the low-frequency oscillation, and the presence of the low-frequency oscillation is identified if the low-frequency oscillation is found at least in the voltage gradient signal and the active power gradient signal or in the voltage gradient signal and the reactive power gradient signal.

6. The method as claimed in claim 1, comprising: detecting a line frequency of the electrical supply network as a further test signal; or transforming the further test signal into a frequency signal as a frequency check signal; temporally deriving or obtaining a difference of the frequency check signal to obtain a frequency gradient signal; checking the frequency gradient signal and at least one further gradient signal for the presence of the low-frequency oscillation, the at least one further gradient signal being one of: a voltage gradient signal, an active power gradient signal or a reactive power gradient signal; and identifying the presence of the low-frequency oscillation in response to the low-frequency oscillation being identified in the frequency gradient signal and in the at least one further gradient signal.

7. The method as claimed in claim 1, wherein detecting the at least one electrical signal of the electrical supply network includes at least one of: detecting a three-phase line voltage; and transforming the three-phase line voltage to produce a direct variable that is a space vector variable of the three-phase line voltage, transforming the three-phase line voltage including determining a positive sequence voltage according to the method of the symmetrical components to form a check signal; or detecting a three-phase feed current; and transforming the three-phase feed current to produce a direct variable that is a space vector variable of the three-phase line voltage, transforming the three-phase feed current including determining a positive sequence current according to the method of the symmetrical components to form the check signal.

8. The method as claimed in claim 1, comprising: identifying a decay of the low-frequency oscillation or a plurality of low-frequency oscillations in response to the gradient signal or respective gradient signals falling below a predetermined termination limit that is less than a check limit of the predetermined criterion.

9. The method as claimed in claim 1, comprising: feeding electrical power into the electrical supply network by a wind power installation or wind farm; and identifying the presence of the low-frequency oscillation by the wind power installation or wind farm.

10. The method as claimed in claim 5, comprising: identifying the low-frequency oscillation in the voltage gradient signal and the active power gradient signal or in the voltage gradient signal and the reactive power gradient signal, and identifying that the low-frequency oscillation in the voltage gradient signal and the active power gradient signal or in the voltage gradient signal and the reactive power gradient signal have the same oscillation frequency or oscillate with the same frequency as the line frequency; and identifying that the low-frequency oscillation is caused in the electrical supply network.

11. The method as claimed in claim 10, comprising: determining the voltage gradient signal, active power gradient signal and reactive power gradient signal as a difference between a maximum value and minimum value in a check period of a corresponding check signal; and determining the same oscillation frequency in the check period, and/or determining that respective time intervals between the maximum value and the minimum value of the corresponding check signal are the same for the voltage gradient signal, the active power gradient signal or the reactive power gradient signal.

12. The method as claimed in claim 1, comprising: determining a gradient quotient as a quotient of two gradient signals; and identifying that the low-frequency oscillation is caused in the electrical supply network depending on the gradient quotient, wherein the gradient quotient is one of: a voltage/active power quotient is formed as a quotient between a voltage gradient signal and an active power gradient signal, or a voltage/reactive power quotient is formed as a quotient between the voltage gradient signal and a reactive power gradient signal.

13. The method as claimed in claim 1, comprising: classifying the low-frequency oscillation; and outputting an oscillation classification of the low-frequency oscillation, wherein the oscillation classification is selected from a list of classifications including: a low-frequency oscillation identified for a line voltage signal and an active power signal; a low-frequency oscillation identified for the line voltage signal and a reactive power signal; a low-frequency oscillation identified for the line voltage signal, the active power signal and the reactive power signal; and a low-frequency oscillation identified for a line frequency and at least one of the line voltage signal; the active power signal and the reactive power signal.

14. A wind power installation or wind farm including a plurality of wind power installations configured to identify low-frequency oscillations in an electrical supply network having a line voltage associated with a nominal line frequency, the wind power installation or wind farm comprising: a controller configured to: detect at least one electrical signal of the electrical supply network as at least one test signal; filter and/or transform the at least one test signal into at least one check signal; temporally derive the at least one check signal or determine a difference between temporally separated values of the at least one check signal to obtain a gradient signal; and identify a presence of a low-frequency oscillation in response to the gradient signal meeting a predetermined criterion.

15. (canceled)

16. The method as claimed in claim 1, wherein the low-frequency oscillations are sub synchronous resonances.

17. The method as claimed in claim 1, wherein the low-frequency oscillation has a frequency between 1 Hz and five times the nominal line frequency.

18. The method as claimed in claim 1, wherein the predetermined criterion is met when the gradient signal exceeds at least one predetermined limit.

19. The method as claimed in claim 1, wherein obtaining the difference of the temporally separated values of the at least one check signal yields a plurality of gradient signals including the gradient signal.

20. The method as claimed in claim 7, wherein the direct variable of the three-phase line voltage or the positive sequence voltage and the direct variable of the three-phase feed current or the positive sequence current are transformed into a voltage signal, an active power signal and a reactive power signal as a voltage check signal, active power check signal or reactive power check signal, respectively.

21. The method as claimed in claim 12, comprising: identifying that the low-frequency oscillation is caused in the electrical supply network in response to the voltage/active power quotient or the voltage/reactive power quotient being negative.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0080] The invention is explained in greater detail hereinafter by way of example using embodiments with reference to the accompanying figures.

[0081] FIG. 1 shows a wind power installation in a perspective representation.

[0082] FIG. 2 shows a wind farm in a schematic representation.

[0083] FIG. 3 schematically shows a flow structure of a method according to an embodiment.

[0084] FIG. 4 shows a schematic diagram of a plurality of check signals.

[0085] FIG. 5 shows a wind energy system with a control device.

DETAILED DESCRIPTION

[0086] FIG. 1 shows a wind power installation 100 with a tower 102 and a nacelle 104.

[0087] A rotor 106 with three rotor blades 108 and a spinner 110 is arranged on the nacelle 104. The rotor 106 is transferred into a rotational movement by the wind during operation and thus drives a generator in the nacelle 104.

[0088] FIG. 2 shows a wind farm 112 with three wind power installations 100 by way of example, which can be identical or different. The three wind power installations 100 are therefore representative of essentially any number of wind power installations of a wind farm 112. The wind power installations 100 provide their power, that is to say in particular the electricity produced via an electrical farm network 114. In this case, the currents or powers produced in each case from the individual wind power installations 100 are added up and a transformer 116 is usually provided which boosts the voltage in the farm, in order to then feed it into the supply network 120 at the feed-in point 118, which is also commonly referred to as PCC. FIG. 2 is merely a simplified representation of a wind farm 112 which does not show any control system, for example, even though a control system is of course present. The farm network 114 can also be configured differently, for example, by a transformer also being present at the output of each wind power installation 100, for example, to mention only one other exemplary embodiment.

[0089] In the flow structure 300, FIG. 3 shows method steps for the method for identifying low-frequency oscillations. A voltage detection block 302 and a current detection block 304 are firstly provided accordingly. The voltage detection block 302 receives the three phase voltages V.sub.1, V.sub.2 and V.sub.3 and transmits a common voltage signal V to a filter block 306. In this case, the three phase voltages V.sub.1, V.sub.2 and V.sub.3 may in particular have been received as a line voltage at a network connection point.

[0090] The current detection block 304 receives the three phase currents I.sub.1, I.sub.2 and I.sub.3 and transmits a common current signal I to the filter block 306. The three phase currents I.sub.1, I.sub.2 and I.sub.3 may in particular have been received as feed currents which a wind energy system has produced and feeds into the electrical supply network at the same network connection point at which the three phase voltages V.sub.1, V.sub.2 and V.sub.3 were also detected.

[0091] The obtained common voltage signal V and the obtained common current signal I are then firstly filtered in the filter block 306. This filtering is adapted to the frequency spectrum of interest. In particular, the filter is designed in such a way that low-frequency oscillations can be maintained as far as possible and are not filtered out.

[0092] In addition, the signals filtered in this way are converted into a voltage effective value V.sub.m, an active power effective value P.sub.m and a reactive power effective value Q.sub.m in the filter block 306. All of these three values are output as signals, i.e., as a voltage signal, active power signal and reactive power signal, wherein each signal reproduces the effective value of the relevant variable depending on the time. These signals output by the filter block 306 can form check signals.

[0093] These three effective value signals are input to the derivative block 308. In the derivative block 308, gradients are determined in each case for the effective value signals by derivation or difference formation and these gradients are compared with a check limit in each case. In this embodiment, the presence of a low-frequency oscillation is assumed if it has been identified that the check limit of the voltage effective value signal V.sub.m is exceeded and also it has been identified that the check limit of at least one of the two remaining effective value signals, namely the active power effective value P.sub.m and the reactive power effective value Q.sub.m is exceeded in each case. Only then is the presence of a low-frequency oscillation assumed. It is of course also possible that all three signals which enter into the derivative block 308 here each exceed their check limit.

[0094] If this check criterion is therefore met, the derivative block 308 outputs a corresponding signal, which is referred to as a trigger signal here. The signal is therefore referred to as a trigger signal because it can be further used in order to trigger responses, i.e., for triggering. Such triggering responses can be the undertaking of damping measures and additionally or alternatively it can be a safety shutdown of the wind energy system which uses this method. It is also possible that the trigger signal is always output but that it has a different value or a different signal amplitude depending on the detected situation, i.e., depending on whether a low-frequency oscillation has been detected.

[0095] FIG. 4 shows in a schematic diagram with measurements recorded over a period of approximately 30 seconds the course of three check signals, namely the voltage check signal V.sub.m, the active power check signal P.sub.m and the reactive power check signal Q.sub.m. These three check signals correspond to the three effective value signals V.sub.m, P.sub.m and Q.sub.m according to FIG. 3, which the filter block 306 outputs there.

[0096] The diagram in FIG. 4 also shows a trigger signal that corresponds to the trigger signal T.sub.rig according to FIG. 3, that the derivative block 308 outputs there.

[0097] The three check signals V.sub.m, P.sub.m and Q.sub.m are represented there in a standardized manner, that is to say standardized to nominal values. In this case, the numbers are represented as “milli”, so that the scale ranges from −1000 to +1000, instead of −1 to +1. According to the proposed method, derivatives are also formed from these three check signals for further evaluation, particularly in the derivative block 308, before a further evaluation takes place. For the sake of simplicity, these derivatives are not represented here.

[0098] It can be recognized in FIG. 4 that all three check signals initially have few oscillations. The voltage check signal V.sub.m and the reactive power check signal Q.sub.m initially have approximately a constant value. Constant reactive power is therefore fed in. The voltage check signal V.sub.m drops slightly, wherein the drop is less than 1%.

[0099] The active power check signal indicates a slightly increasing value. This increase may also be a result of increasing wind speeds. However, the increase in 15 seconds by approximately 3% is comparatively low and in any case does not allow for any conclusion to be drawn regarding low-frequency oscillations.

[0100] Shortly before the time ti, it can be recognized that all three check signals have increasing oscillations. The increase in oscillations appears obvious and easily recognizable in the graph of the schematic representation according to FIG. 4. However, this connection is not easily identifiable for an automatic evaluation by means of a process computer.

[0101] It is therefore proposed to make a derivative of these three check signals in each case, namely the voltage check signal V.sub.m, the active power check signal P.sub.m and the reactive power check signal Q.sub.m. The oscillations then emerge more strongly with such a derivative, which in any case is not represented in FIG. 4. The derivatives then become so large at the time ti that they exceed their respective check limit and the presence of a low-frequency oscillation has therefore been identified.

[0102] This is based on an evaluation which identifies the presence of a low-frequency oscillation if the voltage check signal and the reactive power check signal in each case exceed their check limit, and/or if the voltage check signal and the active power check signal in each case exceed their check limit. In the example in FIG. 4, both criteria are met at the time t1. For the sake of simplicity, the trigger signal T.sub.rig shows at the time ti that this is the case and the trigger signal T.sub.rig jumps from 0 to the value 1. If only one of the criteria is met, the trigger signal T.sub.rig assumes a smaller value, but which is significantly greater than zero, for example 0.8.

[0103] The trigger signal T.sub.rig only assumes the value 0 if none of the criteria is met. The trigger signal T.sub.rig therefore partially drops to this smaller value of approximately 0.8, because the active power check signal or the reactive power check signal have dropped below their check limit there in the temporal ranges. During the entire time represented from the time ti, the voltage check signal has not dropped below its check limit, because in that case the trigger signal T.sub.rig would have dropped to the value 0.

[0104] If it does not assume the value 0, the trigger signal T.sub.rig can then result in a damping measure being initiated, or even a shutdown of a wind energy system taking place, or even the wind energy system being disconnected from the electrical supply network.

[0105] FIG. 5 shows in an illustrative manner a wind power installation 500 with a control device (controller) 502, which is to be regarded as part of the wind power installation 500, precisely like the inverter 504 just shown, and could be arranged in the tower 506 of the wind power installation, wherein the inverter 504 and the control device are represented outside the remaining wind power installation 500 merely for the sake of clarity.

[0106] The inverter 504 receives power produced by the wind as a direct voltage signal and performs the inversion based on this and produces a three-phase feed current I.sub.1,2,3 at a three-phase voltage V.sub.1,2,3. This can be fed into the electrical supply network 510 at a network connection point 512 indicated there via a transformer 508.

[0107] In order to carry out a proposed method for identifying low-frequency oscillations, current and voltage can firstly be measured with an indicated measuring sensor (ammeter, voltmeter and/or multimeter) 514 and be transferred to the detection device 516. The detection device 516 and the measuring sensor 514 can also form a common unit (controller-sensor).

[0108] The detection device 516 therefore detects at least one test signal from the transferred measurements. Voltage and current can each form a test signal here. This test signal or here these test signals are then transferred to the filter unit (filter) 518, which carries out a filtering and in particular carries out this filtering in such a way that substantially only signal components with desired frequencies, that is to say in the range of the expected low-frequency oscillations, remain. These signals filtered in this way are used as check signals and transferred to the derivation unit 520. The symbol of the derivation unit 520 points to a time-continuous derivative, but a discrete derivative by way of difference formation is of course also possible particularly in the case of the presence of discrete signals.

[0109] In any case, the signal derived in this way or these signals derived in this way are transferred to the identification unit 522, which then checks a predetermined check criterion, in particular checks for each received derived check signal whether a predetermined check limit is exceeded in each case. The identification unit 522 can transfer a trigger signal to a process computer 524 as a result.

[0110] In principle, the process computer 524 controls the inverter, takes over further control tasks if applicable, and can also perform this control depending on the trigger signal received by the identification unit 522. Particularly if a low-frequency oscillation or a plurality of low-frequency oscillations have been identified, the process computer 524 can control the inverter 504 in a correspondingly modified manner by specifying a reduction in the power to be fed in, for example. For this purpose, the process computer 524 can additionally perform further controls in the wind power installation, something which is not shown in FIG. 5, such as adjusting the rotor blades, for example, in order to also extract less power from the wind accordingly.

[0111] Additionally or alternatively, it is possible that in the event of identified low-frequency oscillations, for protecting the wind power installation, feeding is suspended and, if applicable, a safety switch is opened, which in any case is not shown in FIG. 5 for the sake of simplicity.

[0112] In particular, a method for identifying low-frequency oscillations has therefore been proposed. This takes into account that energy systems are oscillatory systems which possess natural modes below and also possibly above the system frequency. The line frequency is to be assumed here as a system frequency, i.e., usually 50 Hz or 60 Hz. Oscillations of this type can affect system stability when excited.

[0113] Wind power installations can also contribute to the stabilization, or even to the destabilization in the case of incorrect handling, of the energy system. It should be noted that the lifespan of a wind power installation can be approximately 25 years, and the energy system can also change and develop significantly during this time.

[0114] For the observation of low-frequency oscillations now proposed, which oscillations are also referred to as energy system oscillations or power system oscillations (PSO), not only a warning system for the operation of wind power installations or wind farms can be used, but this information as a result of the observation can also be used, in order to use suitable damping signals through the wind farm or possibly through a wind power installation for damping the energy system oscillations.

[0115] The proposed method can also be implemented as an algorithm in a control device, in particular a process computer. In the case of a wind farm as a wind energy system, a central farm computer or a central farm control unit can also be considered for this purpose, on which the method can be implemented. In particular the detection device, filter unit, derivation unit and identification unit, as illustrated in FIG. 5, can also be combined in a common control unit (controller), or they can be implemented as algorithms or software blocks.

[0116] The proposed algorithm or the proposed method is particularly based on the analysis of voltage and power gradients. Moreover, evaluations can be performed on site at the wind power installation, or in the wind farm, or even in a remote monitoring center. There is then the possibility that the required data is transmitted via SCADA for this purpose.

[0117] Moreover, the proposed method can also be applied for consumer units, and in principle also for conventional power stations. For example, in the event of identified low-frequency oscillations, consumer units can optionally change their behavior or optionally disconnect from the electrical supply network.

[0118] In particular, a solution is also provided which enables a method for online detection of energy system oscillations.

[0119] In particular, the method makes provision for an online analysis of transient measurement data at a network connection point of a wind farm. This is particularly advantageous because a central farm control unit is usually arranged there in wind farms. The analysis can therefore be performed immediately and on site. It is proposed that voltages and currents are evaluated, filtered in a suitable manner and that finally the calculation of the temporal gradients of the filtered voltage and calculated power signals takes place. Low-frequency oscillations can then be detected as a result of suitable parameterization of the check limits, which can also be referred to as a threshold value.