EP3776787A1

EP3776787A1 - Method for feeding electrical power into an electrical supply network

Info

Publication number: EP3776787A1
Application number: EP19715087.3A
Authority: EP
Inventors: Johannes BROMBACH
Original assignee: Wobben Properties GmbH
Current assignee: Wobben Properties GmbH
Priority date: 2018-04-05
Filing date: 2019-04-02
Publication date: 2021-02-17
Also published as: CA3095344C; US11527890B2; DE102018108023A1; WO2019192993A1; US20210151985A1; CA3095344A1; CN111989840A

Abstract

The invention relates to a method for feeding electrical power into an electrical supply network by means of a decentralised feeding unit, in particular by means of at least one wind turbine, in which the feeding unit is connected to a network node point for feeding in the electrical power, the network node point is connected directly or via a connection to a transformer point for transmitting the electrical power from the network node point to the transformer point, optionally via the connection, the transformer point is connected via a transformer to a network segment for transmitting the electrical power from the transformer point via the transformer to the network segment. The method comprises the steps of feeding in electrical active power into the electrical power supply at the network node point, feeding in electrical reactive power into the electrical power supply at the network node point, detecting a change to be made to the active power to be fed in, changing the active power fed in according to the detected change to be made, and limiting a temporal change of the reactive power fed in as active power fed in is changed and/or immediately thereafter, in order to counteract a voltage increase at the transformer point and/or in the network segment, and/or temporarily activating a voltage control as a function of the changing of the active power fed in, in order to carry out a voltage control at a reference point, in particular at the network node point, during and/or immediately after the changing of the active power fed in, in order to adjust the voltage dynamically at the reference point or network node point, and/or to conduct the voltage along a trajectory, in particular along a ramp.

Description

Method for feeding electrical power into an electrical supply network

The present invention relates to a method for feeding electrical power into an electrical supply network by means of a decentralized feed unit, in particular by means of at least one wind turbine or a wind farm. Furthermore, the present invention relates to a corresponding wind turbine, which can perform such a method, and the invention also relates to a wind park with multiple wind turbines, the wind park can perform such a method. In particular, the present invention also relates to a wind park that can perform such a method of feeding by means of its wind turbines.

Wind turbines, wind parks or other decentralized feed-in units can nowadays not only feed into the electrical supply network in grid parallel operation, but they can also make a contribution to controlling and stabilizing the electrical supply network. A fundamental situation, which falls under the fact that in the case of an excess supply such decentralized feed-in units have to reduce their fed-in power to a corresponding requirement, which may, for example, come from the grid operator. In particular, wind turbines, wind parks or other decentralized feed-in units that feed by means of an inverter or inverter technology, their power very quickly, namely within a few seconds, e.g. 5 seconds or less.

Such a fast control behavior or such a fast control capability can be very advantageous for controlling the electrical supply network, but it can also undesirable effects occur. It is often the case that such a decentralized feed-in unit feeds with an at least temporarily predetermined cos (cp) which is less than one. It is also fed reactive power. The decentralized feed unit thus feeds the desired power by feeding in a corresponding current, the current having a corresponding phase angle other than 0 with respect to the phase position of the mains voltage, that is to say the voltage in the electrical supply network. This reactive current or the corresponding reactive power is required to set a voltage level. In particular, there is a need for this long connection connections between a network connection point, is fed into the, and a transformer, which transforms the voltage to a higher voltage level. But there are also other constellations into consideration, in which the injected reactive power or the fed reactive current affects the voltage level.

Now, if the described power reduction is performed, so the injected current is reduced. If the phase angle is unchanged, the cos (cp) also remains unchanged, and thus the reactive power is also reduced by reducing the injected current. This, in turn, can lead to an undesirable voltage change both at the grid connection point and especially at the transformer, in particular at a substation, although ideally this should not be so, because the cos (cp) should be set to cause a voltage drop across a line is compensated. However, such always suitable compensation presupposes a linear system and usually only works optimally at one point. In principle, voltage changes, especially at the substation, can be compensated by a corresponding step transformer. The tapped transformer then changes the effective winding ratio of primary to secondary side or vice versa, in particular by mechanically changing a corresponding tap, and thus changes its transmission behavior. A voltage change can thus be compensated by appropriate adjustment of the tapped transformer. However, known step transformers are usually not able to compensate for such a rapid voltage change as can be achieved by the rapid change of the fed power by means of the decentralized feed unit. Thus, due to the lack of reactive power, there is a short-term increase in the voltage on the primary side, which can not be corrected immediately and thus can occur especially as a voltage spike in the electrical supply network.

To counteract this, it is known to artificially slow down the control behavior of such decentralized feed-in units. Frequently, the desired power reduction is known at an early stage because, for example, it provides a reduction in the operation of a wind energy plant at a certain time, for noise reduction. If, for example, for reasons of sound reduction at eight o'clock in the evening the operation of a wind energy plant is to be reduced, this can already be started at five minutes before eight or ten minutes before eight, when the power is then slowly reduced by means of a corresponding ramp at this earlier time so that the power actually reached the reduced value at eight o'clock in the evening. But such a measure leads to a loss of revenue, because even earlier than necessary, a reduction in performance is made.

The German Patent and Trademark Office has in the priority application for the present PCT application the following state of the art research: DE 10 2012 212 366 A1; US 2013/0168963 A1 and US 2014/0375053 A1

The present invention is therefore based on the object, at least one of the o.g. To address problems. In particular, a solution should be proposed in which a decentralized feed units can be reduced in power with as little loss of yield as possible. At least should be proposed to previously known methods an alternative.

According to the invention, a method according to claim 1 is proposed. This method thus relates to the feeding of electrical power into an electrical supply network by means of a decentralized feed unit. The decentralized feed unit can in particular be a wind turbine or several wind turbines, in particular it can form a wind park with several wind turbines. The feed unit is connected to a grid connection point for feeding the electrical power. The grid connection point is in turn connected via a connection connection to a transformer point, namely for transmitting the electrical power from the grid connection point via the connection connection to the transformer. The transformer point is in turn connected via a transformer to a network section for transmitting the electrical power from the transformer point via the transformer to the network section. Accordingly, this topology thus includes the feed-in unit, the network connection point, the connection connection and the transformer, which is finally connected to a network section. Presupposing this topology, active electrical power is fed into the electrical supply network at the grid connection point. In addition, reactive electric power at the grid connection point is fed into the electric grid. Furthermore, a change to be made to be fed active power is detected. It is therefore detected, for example, when the active power is to be increased or decreased in the near future or at a certain point in time. Accordingly, the injected active power is changed.

For this purpose, a limiting of a temporal change of the fed-in reactive power when changing the injected active power and additionally or alternatively proposed immediately afterwards. This is proposed to counteract a voltage increase at the transformer point and / or in the network section.

In this case, a temporal change of the fed-in reactive power is deliberately proposed and this can in particular mean the specification of a ramp, that is to say a gradient, or of a limit gradient, which determines how fast the reactive power is maximally changed. However, such a limitation is not provided for the active power. The active power can therefore be reduced immediately at the time when it has to be reduced. It does not need to be reduced before. For the limitation of the temporal change of the fed-in reactive power, this is also proposed for this time or starting at this time, starting immediately or immediately thereafter.

Likewise, this can also be provided for an increase in active power. At the time when the active power is to be increased or especially to which it may be increased, it is then immediately increased to the new value, so that immediately as much active power as possible and allowed to be fed. It is thereby avoided that the active power slowed down from the same time is ramped up with a ramp, because this would yield again yield.

In addition, this also makes a comparatively fast and easy implementation possible, because in this variant, especially for the reduction of active power does not have to be planned in advance.

Preferably, the proposed limitation is performed only temporarily, namely only for the event of reducing the active power feed.

The method for limiting the reactive power gradients is particularly advantageous when the wind farm is connected directly to the UW and is in a cos (cp) control. Here it was particularly recognized that different ratios of resistance to reactance between sections in the electrical supply network on the one hand and sections in a transformer on the other hand can be expected. This makes it difficult to keep tensions constant in both sections. Here it is also taken into account that cos (cp) can be well suited for voltage regulation for lines, whereas a voltage-dependent reactive power regulation (Q (U)) on the transformer, ie especially on a substation, and in the electrical supply network are often better suited. It has also been recognized that it must be taken into account that in some arrangements a voltage-dependent reactive power gelung is not provided or is only complicated to implement. Then, if necessary, a voltage regulation must be carried out using cos (cp). Alternatively, instead of limiting the time change of the fed-in reactive power, a temporary activation of a voltage regulation as a function of changing the injected active power is proposed. Preferably, this temporary activation is provided for a period of less than 10 minutes. This temporary voltage regulation works in such a way that during and / or immediately after changing the injected active power, voltage regulation is carried out at a reference point, in particular at the network connection point, in order to dynamically correct the voltage at the reference point or network connection point and / or along a trajectory , especially along a ramp. Whether to control a voltage at the grid connection point or at another reference point, such as the transformer point, may also depend on a default by the grid operator.

It is thus proposed according to this variant, a voltage regulation, in which the voltage can be controlled in particular via a reactive power setting. Controlling a voltage by adjusting a reactive power is basically known, but here it is proposed to apply such a regulation only temporarily. Only for the short event of active power reduction, this voltage regulation is proposed. In particular, there may otherwise be a cos (cp) control that is suspended for the event of changing the active power.

In particular, the temporarily activated voltage regulation is active only when changing the injected active power and / or immediately thereafter. Voltage regulation operates to cause the voltage at the grid junction to be wholly or partially returned to a value that the voltage at the grid node immediately prior to changing the injected active power had changed from a value that is changed by changing the injected active power. Here it is thus particularly suggested to reduce the tension to the old value. This should be done dynamically to avoid signal peaks. For example. For this, a vibration-free behavior of a delay element of second order can be specified. Preferably, the behavior of how the voltage is conducted, namely to be particularly attributed, is fixed by a trajectory. Thus, a voltage curve is fixed and the behavior can be kotrolliert. Signal peaks are avoided.

Preferably, the temporarily activated voltage regulation is carried out in particular when the network connection point is directly connected to a transformer point is connected, so if there is no connection connection or the connection to the reactive power voltage behavior has no or no significant influence. For this purpose, it has been particularly recognized that there is no influence by a connection connection and therefore an immediate voltage regulation to the reference point can be advantageous.

The limiting of the time-dependent change of the fed-in reactive power preferably takes place in such a way that the fed-in reactive power is changed according to a change function, in particular according to a ramp with a slope limited in magnitude. It is thus given a ramp for changing the reactive power, thereby avoiding a reactive power jump. The steepness of this ramp and thus the speed of the change of the reactive power is preferably given as a limit, so that it is not exceeded in any case. A lower slope may be considered.

In addition or as an alternative, the limiting takes place in such a way that a radar frequency is specified for the injected reactive power which predetermines a maximum temporal change of the supplied reactive power according to the amount, in particular such that the fed-in reactive power with a temporal Gradients is changed, which does not exceed the magnitude of the gradient limit. As a result, a maximum rate of change, which can not be exceeded, is predetermined by such a gradient limit value. A weaker change is allowed. As a result, the change in the reactive power is basically arbitrary, but is limited in terms of excessive changes in time.

According to one embodiment, it is proposed that the transformer has a primary side to which the transformer point is connected, that the transformer has a secondary side to which the network section is connected, and that the transformer is designed as a step transformer and a transmission ratio from the primary side to Secondary side can be set, thereby controlling a voltage level at the transformer point. Thus, for this embodiment is based on such a step transformer and it is proposed that the limitation of the change over time of the injected reactive power is such that a resulting from the change in the fed reactive power voltage change at the transformer point or on the primary side is so slow that the tapped transformer she can control. Here it was particularly recognized that the reactive power can also influence the voltage level on the primary side of the transformer. Although a step transformer used there can compensate for a voltage change, so that a voltage change on the primary side does not lead, or reduces, to a corresponding voltage change on the secondary side, but this can only be done comparatively slowly.

Thus, it was recognized to adapt the limitation of the temporal change of the input reactive power to the dynamics of the tapped transformer. Particularly by means of a correspondingly flat ramp for controlling the change in the reactive power, it can accordingly be achieved that the voltage on the primary side of the step transformer also changes only slowly. In this case, it is basically proposed to provide the limitation of the time change of the fed-in reactive power only if a change to be made in the active power to be fed has been detected, that is to say if intentional gradients of the active power or reactive power are present. The limitation is therefore only temporary and should not limit any changes.

Preferably, it is provided here that a ramp function is specified for the change of the reactive power, which predefines a change from a reactive power initial value to a reactive power end value and for this provides at least a duration of one minute, in particular of at least two minutes and preferably of at least five minutes. Thus, the rate of change of the reactive power is significantly limited and basically adapted to a dynamic of the tapped transformer. It should be noted that a change in reactive power, if not limited as proposed, can be accomplished in a few seconds or even faster by a feed unit using an inverter. It is therefore proposed a slowdown many times over the technical possibilities of an inverter.

In particular, it is proposed that despite limitation of the time change of the fed reactive power for the fed active power no limitation of a temporal change is provided. It is therefore expressly proposed to provide different dynamics for the active power and the reactive power, namely for the reactive power a very slow dynamics and for the active power a fast or unlimited dynamics. The change in the active power is then only technically conditioned by the decentralized feed-in unit and in particular by its inverter. But even a change in power extraction from the wind in a wind turbine can in modern pitch-controlled wind turbines within be carried out for a few seconds. Such a speed of change of the active power should also be allowed here, while the reactive power is limited in its temporal change, namely, in particular is very much limited compared to the said rate of change of active power. According to one embodiment, it is proposed that when changing the fed active power and / or immediately thereafter a voltage regulation at the network connection point or at another point in the network, in particular the secondary side of the transformer is performed to the voltage at this point, thus also as a control point may be designated to balance to a value, or to guide along a trajectory, in particular along a ramp.

It is thus proposed in particular that the change in the reactive power is temporarily incorporated into a control loop. This control loop detects the voltage at the network connection point or another point or control point in the network, which could also be done indirectly via a measurement filter or state observer, but preferably directly via a measurement, and returns this so detected voltage at the network connection point for the scheme. In particular, a desired actual value comparison is then carried out, in which a desired voltage value is compared with an actual voltage value, namely with the detected voltage, and this control deviation is then used as an input variable for reactive power regulation. The reactive power is thus adjusted as a function of such a voltage deviation. This can also be used to control the voltage along a trajectory, namely by specifying instead of a constant voltage value a voltage curve as a voltage setpoint and with this the described comparison with the detected voltage, ie with the actual voltage, is then performed. In particular, by such a guidance of the voltage along a trajectory a desired voltage profile can be predetermined and this can also be done so that the voltage waveform leads to a voltage change not only at the grid connection point, but also at the primary side of the transformer, ie at the T ransformatorpunkt, the Can regulate tapped transformer. This avoids excessive voltage changes through targeted regulation.

According to one embodiment, it is provided that a power factor control is provided for the input of electrical reactive power, which can also be referred to as cos (cp) control. In such power factor control, the injected reactive power is adjusted depending on the injected active power so that a given power factor, which is also referred to as cos (cp) or coscp results.

In this case, the fed reactive power is initially changed simultaneously with the change in the active power fed so that the power factor remains unchanged. However, a new value for the power factor as a function of a voltage which changes as a result of the change in active power and reactive power is then specified at the network connection point. The new value for the power factor can, however, also be predefined as a function of a changing voltage at the network connection point which is to be expected as a consequence of the change in the active power and reactive power. In other words, a change in the active power and reactive power leads to a change in the voltage. This can be counteracted by changing the power factor. This can either be done so that the immediately resulting voltage change is converted as soon as possible into a modified power factor, thereby counteracting the voltage, or this voltage change can already be anticipated.

Due to the topology or previous measurements, it may be well known how the voltage at the grid junction will change in response to the change in real power and reactive power. Insofar as such a change in the voltage is not desired, this can be counteracted by a corresponding change in the power factor, in particular in such a way that the power factor is changed right at the beginning of the change of the active power, so that the reactive power does not change so much that an undesirably high voltage change occurs at the network connection point or another point or control point in the network.

In any case, in both cases, the result will be a modified power factor, which is changed from the described initial situation. It is now proposed to return this changed power factor to the former power factor, namely the power factor then and still provided. The power factor has therefore been changed only temporarily to counteract the change in the voltage at the network connection point, or another point or control point in the network. The power factor is thus fed back again, wherein the feedback is delayed and also or alternatively carried out via a time function.

It is also proposed here that the power factor is not due to a step function, which can also be referred to as a step function, but gradually. This can take place via different delays, such as, for example, via a PT1 behavior, to name only one variant. But it can also be speaking time function are returned and a simple time function is a corresponding ramp, which starts at the temporarily changed power factor and ends at the previous, ie predetermined power factor. Instead of a ramp but also other functions come into consideration, which should as possible but have no or at most small jumps.

This is thus based on a solution for a conventional feed using a power factor control. Such a power factor control basically provides to maintain a fixed power factor and always change the reactive power according to this power factor with active power changes. For this control concept, it is proposed to keep it as unchanged as possible, but to temporarily change the power factor for the described critical case of the undesired strong change in the voltage at the network connection point. Thus, only the described power factor change needs to be made and the overall implemented power factor control can be maintained.

According to one embodiment, it is proposed that

for the supply of electrical reactive power, a power factor control (cos (cp) control) is provided, in which the reactive power fed in response to the injected active power is set so that a predetermined power factor (cos (cp)) results,

when changing the active power fed it is changed along a ramp or trajectory. This change is carried out in particular in a time of less than one minute, in particular in the range of 10 seconds to 40 seconds.

It is further suggested that

when changing the applied active power along the ramp or trajectory, a voltage regulation is used to keep the voltage at the reference point constant, at least to counteract a voltage change by changing the applied active power, the voltage regulation for this purpose sets the reactive power and the power factor control (cos ( cp) control) is deactivated during this voltage regulation,

after the change in the injected active power, a reactive power value resulting from the voltage regulation is fed to a new reactive power value, which would result from the deactivated power factor control (cos (cp) control) the reactive power value is guided in particular via a ramp and / or trajectory to the new reactive power value. For this feedback, a period of 2 minutes to 10 minutes is preferably provided, in particular a period of 3 to 7 minutes. Subsequently, the deactivated power factor control (cos (cp) control) is reactivated, namely as soon as the reactive power value reaches the new reactive power value has reached.

Accordingly, a regular feed with active power supply (P) and reactive power setting (Q) is also initially assumed in accordance with a set power factor cos (cp). Then there is an active power ramp with temporarily activated voltage regulation. This results in the end of a modified Q-value, which is chosen so or adjusts itself by the voltage control that, if possible, the voltage is held substantially at the old value.

Then, the difference between a reactive power setpoint (Q setpoint) corresponding to the power factor control (cos (cp control) and the actual reactive power value (Q value) corresponding to the voltage regulation is determined. This is followed by a slow ramp-down of the reactive power value (Q value) to the value in accordance with the power factor control (cos (cp) control). This is done so that there is only a continuous change in voltage. A tap changer that is a variable transformer can then react to this continuous change in voltage.

Preferably, it is proposed that the changing of the injected active power is carried out via an active power jump, although, on the other hand, the changing of the reactive power is performed via a time function, in particular via a ramp function. It is thereby achieved in particular that an active power change can be carried out as quickly as possible in order to avoid power losses. Unwanted voltage jumps or otherwise unwanted voltage changes, especially at the grid connection point or other point in the electrical supply network, but can be avoided by the non-jump adjustment of reactive power. The choice of the point at which unwanted voltage jumps or otherwise undesirable changes in voltage can be avoided may depend on a network operator, for example by pretending to do so.

Preferably, the limitation of the change of the input reactive power depends on the type and / or size of the connection connection. Here, it was particularly recognized that the connection, in particular a line between the network link tion point and the transformer point, ie the transformer, can vary greatly in their impedance. There may be topologies with large and small impedances, but in particular the reactance, ie the reactive component of the impedance, can be very different. Depending on the type of cable, the reactance, ie the reactive component, can be as large as the resistive resistor, ie the effective component, the impedance, or many times greater than the ohmic resistance. This has particular effects on reactive power-dependent voltages or on how reactive power changes change the voltage. This is preferably taken into account when limiting the time change of the fed-in reactive power. According to the invention, a wind turbine is also proposed for feeding electrical power into an electrical supply network, wherein

the wind turbine is connected to a grid connection point for feeding the electrical power,

the network connection point is connected via a connection connection to a transformer point, for transmitting the electric power from the network connection point via the connection connection to the transformer point, the transformer point is connected via a transformer to a network section, for transmitting the electrical power from the transformer point via the transformer to the network section, and the wind turbine comprises

an inverter for injecting active electrical power into the electrical supply network at the grid connection point,

an inverter controller for controlling a supply of reactive electric power to the electric utility network at the grid connection point; an input interface for detecting a change in the active power to be fed in, wherein

the inverter controller is prepared to vary the injected active power in accordance with the detected change to be made, and wherein the inverter controller is prepared to limit a time change of the reactive power input when changing the inputted one

Active power and / or immediately thereafter to counteract a voltage increase at the transformer point and / or in the network section, and / or the inverter control is prepared to perform a temporary activation of a voltage regulation as a function of changing the injected active power in order to or immediately after changing the injected active power to perform a voltage control at a reference point, in particular at the network connection point to the voltage at Dynamically align reference point or network connection point, and / or to lead along a trajectory, in particular along a ramp.

In particular, the feeding is carried out by means of an inverter, which receives its power from a generator. The inverter can be controlled via the inverter controller and can also be used to control other process steps. In particular, such an inverter controller can control the changing of the injected active power according to the detected change to be made, and it can also limit the time variation of the supplied reactive power. In particular, the inverter controller can implement corresponding control control algorithms for each and generate resulting setpoint values for the active power, reactive power and / or the power factor or phase angle. These setpoints can then be converted by the inverter accordingly. In this case, a desired value for a current to be fed can also be specified.

Preferably, the wind turbine is prepared to perform a method according to at least one of the embodiments described above. For this purpose, corresponding algorithms can be implemented in the inverter control or another control unit, preferably at least one voltage sensor is provided to detect a voltage at the network connection point or another point in the network. According to the invention, moreover, a wind park comprising several wind turbines is proposed which comprises at least one wind energy plant according to the invention and / or at least one method according to one of the embodiments described above, which relate to a method for feeding. For this purpose, the corresponding method can be implemented in a central control unit of the wind farm or is executed individually by the wind energy installation.

According to the invention, a feed arrangement for feeding electrical power into an electrical supply network is proposed, comprising

a decentralized feed-in unit, in particular a wind turbine, a grid connection point,

- A transformer with a transformer point, where

the supply unit is connected to the grid connection point for feeding the electrical power,

the network connection point is connected directly or via the connection connection to the transformer point, for transmitting the electrical power from the network connection point via the connection to the

Transformer point,

the transformer point is connected to a network section via the transformer, for transferring the electrical power from the transformer point via the transformer to the network section, and the feeder arrangement comprises an inverter for feeding active electrical power into the electrical supply network at the network connection point,

the inverter controller is prepared to change the injected active power according to the detected change to be made, and wherein

- The inverter control is prepared to limit a time

Modification of the fed reactive power when changing the injected active power and / or immediately thereafter to counteract a voltage increase at T ransformatorpunkt and / or in the network section and / or the inverter control is prepared to a temporary activation of a voltage regulation in response to varying the injected active power to perform during and / or immediately after changing the injected active power, a voltage control at a reference point, in particular at the network connection point to dynamically adjust the voltage at the reference point or network connection point, and / or to lead along a trajectory, in particular along a ramp.

The feed arrangement thus has at least one decentralized feed-in unit, in particular a wind energy plant, and additionally also a grid connection point, a connection connection a transformer with a transformer point. Otherwise, their design and operation according to the embodiments described above.

The feed arrangement thus preferably has a wind power plant according to at least one embodiment of a wind energy plant described above. Additionally or alternatively, it has a wind park according to at least one embodiment of a wind farm described above. Additionally or alternatively, the feed arrangement is prepared to carry out a method according to at least one embodiment of a method described above. The effect and the advantages arise accordingly from the respective explanations of the respective embodiments mutatis mutandis.

The invention will now be described by way of example with reference to embodiments with reference to the accompanying drawings. Figure 1 shows a wind turbine in a perspective view.

FIG. 2 shows a wind park in a schematic representation. FIG. 3 shows a feed arrangement in a schematic representation. FIG. 4 shows an equivalent circuit diagram for the feed arrangement of FIG. 3. FIG. 5 shows a diagram of a feed change according to the prior art.

FIG. 6 shows a further diagram of a feed change. Figure 7 shows a diagram of a feed change according to a proposed

Embodiment.

FIG. 8 shows a diagram of a feed change according to a further embodiment.

FIG. 9 shows a diagram of a feed change according to yet another

Embodiment.

FIG. 1 shows a wind energy plant 100 with a tower 102 and a nacelle 104. A rotor 106 with three rotor blades 108 and a spinner 110 is arranged on the nacelle 104. The rotor 106 is set in rotation by the wind in rotation and thereby drives a generator in the nacelle 104 at. For feeding in an inverter 130 is provided, which can be controlled by an inverter controller 132. The inverter control is additionally provided with an input interface 134, via which setpoint values, in particular an active power setpoint Ps, can be input. FIG. 2 shows a wind farm 112 with, by way of example, three wind turbines 100, which may be the same or different. The three wind turbines 100 are thus representative of virtually any number of wind turbines of a wind farm 112. The wind turbines 100 provide their power, namely, in particular, the power generated via an electric parking network 114 ready. In this case, the respectively generated currents or powers of the individual wind turbines 100 are added up and usually a transformer 116 is provided, which transforms the voltage in the park, in order then to feed into the supply network 120 at the feed point 118, which is also generally referred to as PCC. There is also provided a tapped transformer 124 which is connected via a terminal connection 122 to the feed-in point 18, which may also be referred to as a grid connection point. Via the step transformer 124, the voltage is further transformed up to the voltage in the supply network 120. The wind turbines 100 may be formed as shown in FIG. 1, including inverter 130 and inverter controller 132 with input interface 134. However, a central controller for the park 112 may also be provided.

Fig. 2 is only a simplified representation of a wind parks 1 12, for example, shows no control, although of course there is a controller. Also, for example, the parking network 114 can be designed differently, for example by also a transformer at the output of each wind turbine 100 is present, to name just another embodiment.

The feed arrangement 300 of FIG. 3 has a wind park 302 as a decentralized feed unit. This wind park 302 is connected to a network connection point 304 and connected via a connecting cable 306, which forms a connection here, via a busbar 308 to a transformer point 310. The bus bar 308 may also be considered part of the transformer point 310 because the bus bars 308 basically form the connection hardware of the transformer point 310. Thus, the connection cable 306 is connected to the transformer point 310. Here, the connection cable 306 is designed here by way of example to a voltage level of 20 kV, and this voltage can be transformed via a transformer 312 to a higher voltage, which here is for example 1 10 kV and forms the voltage level of the network section 314. Accordingly, the transformer 312 is connected to the network section 314. The transformer has here a primary side 316 and a secondary side 318. Illustratively, in parallel with the interconnect cable 306, a parallel string 320 is shown illustratively, which may also connect the mesh interconnection point 304 to the busbar 308 and the transformer point 310, respectively, when a disconnect switch 322, which is also illustrative, is closed. Also illustratively, various consumers, namely industrial consumers 324 and non-industrial consumers 326 in the region of the connection cable 306 or the parallel strand 320 are indicated, which can be connected there in each case. Despite the same reference numerals 324 and 326, these consumers may still be different. For this feeder arrangement 300 shown, a method is now provided in which the wind park as a representative decentralized feed unit feeds both active power and reactive power at the grid connection point 304. The fed active power can then be changed according to a corresponding specification, ie a change to be made. For example. If required, it can be halved. At the same time a temporal change of the fed reactive power is limited. This is done so that a voltage increase at the transformer 312, namely at the transformer point 310 and / or at the network connection point 304 is counteracted. This does not have to mean that a voltage increase is excluded here, but it is at least reduced compared to a voltage increase in magnitude, which would result without such a temporal change of the input reactive power.

In this way, it can be achieved that voltage jumps at the transformer and / or at the network connection point are limited in the case of an unfavorable or even incorrect parameterization of a cos (cp) -control or -control. Such an unfavorable parameterization may, for example, occur depending on a reactive power budget, for example if a reactive power feed or effect is misjudged, or reactive power can not be supplied as intended.

In the case of correct, optimal or ideal parameterization, the cos (cp) at the network connection point would be set such that the voltage would scarcely change given a change in power. In the event that a wind park is connected directly to a busbar such as busbar 310, ideally a purely inductive character would be assumed and cos (cp) parameterized to one. However, there are regular deviations from this ideal state, which are taken into account here. FIG. 4 now shows an equivalent circuit diagram by way of example for the feed arrangement 300 of FIG. 3 or for a part thereof. The equivalent circuit diagram of FIG. 4 is fundamentally initially based on the fact that the disconnect switch 322 is open as shown in FIG. If it were closed, but would give a very similar equivalent circuit diagram, but with other concrete values, namely with other concrete impedances or at least one other impedance.

Accordingly, FIG. 4 initially shows the wind park 302 again, which feeds into the network connection point 304. The connection section to the busbar 308 and thus to the transformer point 310 has a connection impedance 406. This connection impedance 406 is designated as impedance Z3. Here very concrete values are given, to which also the results of the diagrams shown below refer. However, these values mentioned and explained below are given by way of example only and may have both small and large deviations therefrom. In any case, the exemplary connection impedance 406 has an ohmic component, that is to say active component of 5.78W, and a reactive resistance value, that is, a reactive component of 4.83W. As already explained, the exact values do not matter, but for the connection impedance 406 it can be stated that it has approximately the same active and reactive component here. An active current and a reactive current in each case the same amplitude would thus lead to this connection impedance 406 about the same voltage drops.

The busbar 308 or the transformer point is followed by a transformer impedance 412 and a network impedance 414 of a higher-level network. The network or the network section 314 can be transformed to the primary side 316 of the transformer 312 as a voltage source 415 with the voltage level 20 kV. Both the transformer impedance 412 and the network impedance 414 are thus impedances which are transformed to the low-voltage side, that is to say the voltage side of the primary side 316 of the transformer 312. In this case, the transformer impedance 412 correspondingly represents a transformed impedance of the transformer. The network impedance 414 represents a transformed impedance of the network section 314.

Here, too, exemplary values are given, which were taken into account below, but which may also be different. In any case, the values of the transformer impedance 412 are 0.14W ίϋt the ohmic component and 3W ίϋt the component of the reactance. In the case of the network impedance 414, there is an ohmic component of 0.07 W and a component of the Reactive resistance of 0.4 W before. It can be seen that especially the transformer impedance 412 has a much greater reactance value than the value of the ohmic resistance. Here is about a factor of 20 between the two shares. It can thus be seen that a change in the reactive current through this transformer impedance 412 results in a substantially higher voltage change than a change in the active current. Accordingly, an undesirable voltage change may particularly, but not exclusively, occur at the transformer impedance 412 when active and reactive power supplied by the wind park 302 are varied to the same extent. In principle, the injection of active and reactive power, which also includes the negative feed of reactive power, can be selected so that - clearly speaking - the reactive power counteracts a voltage increase by the active power. This applies in any case to the network connection point if a connection connection dominating here, ie a dominant cable impedance of the connection connection, is present. But if both values are changed equally, this can lead to a new situation at another point where this compensation no longer exists. This was particularly recognized as a problem in wind parks, in which the network connection point are connected directly or with little dominant connection connection with a transformer, in particular substation. This is especially true if the grid connection point is directly connected to the transformer, or even in the substation and if the wind parks are in a cos (cp) control. This may, for example, be the case if a corresponding control of the reactive power budget requires this. It was recognized that in such a case, the active current hardly acts lift, but the reactive current has a voltage lowering effect. If, in such a case, the reactive current is missing, there is a sudden increase in the voltage during the power increase.

The following diagrams illustrate different ways to change the active and reactive power and show the resulting potential effects. FIG. 5 shows a diagram showing a known type of variation.

In the diagram of FIG. 5, power and voltage characteristics are plotted over time. It is initially assumed there from a stable feed situation until the time tp an active power reduction specified and then implemented. In that regard, the diagram shows a fed PLEX Parkwirkleistung that there is reduced to about 3MW, for example, at the time of reduction te by example 10MW. The reduction basically takes place immediately, so that the diagram of FIG. 5 shows approximately a stepped progression for the reduction of the active power PWEA. Due to a fixed predetermined phase angle, the injected reactive power QWEA is also reduced at the time of reduction fp, namely basically stepwise. The diagram then shows deviations between the active power PWEA and the reactive power QWEA, which, however, are partly also relevant to the depiction of connections and are of less importance here. In any case, the injected reactive power QWEA together with the active power is also reduced quickly. As a result or related consequences, both the voltage UNVP at the network connection point 304 and the voltage Uss at the busbar 308, which in turn corresponds to the voltage at the transformer point 310, may change.

It can be seen that the reduction of the active power PWEA and the reactive power QWEA leads to a reduction of the voltage UNVP at the network connection point. This is particularly related to the fact that the connection impedance 406 has a higher ohmic component compared to the reactive power component. This results in a reduction of the active power to a voltage reduction, which counteracts the simultaneous reduction of reactive power. However, since the reactance component of the connection impedance 406 is smaller than the resistive component, the simultaneous reduction of reactive power can not completely counteract the voltage reduction due to the reduction in active power. At a phase angle optimally set for this network link point, however, the voltage change at the network link point (NVP) would be minimal. The voltage U.sub.ss at the busbar is somewhat different because, in particular, the transformer impedance 412 but also, albeit somewhat less, the network impedance or transformed network impedance 414 each have a significantly higher reactance component than the ohmic component. There, the voltage-boosting effect due to the reactive power reduction is stronger than the voltage-lowering effect due to the reduction in active power. For this reason, the voltage at the busbar Uss undesirably increases at the time of reduction t _R.

A particular problem here is that this voltage increase is accompanied so rapidly, namely directly with the active power reduction and reactive power reduction, that a tapped transformer can not compensate this fast enough. Especially The transformer 312 may be designed here as a step transformer to counteract a slow increase in voltage at T ransformatorpunkt 310.

The first countermeasure proposed in the prior art is to reduce the active power PWEA slowly, so that the reactive power GWEA is also reduced slowly. This is shown in FIG.

From FIG. 6, it can be seen in the diagram, however, that to achieve a target value for the active power PWEA to be reduced, which is to be achieved at the time of reduction tp, the reduction must begin earlier. For this, the reduction of the active power is already started at the preliminary time point tv. In fact, this ensures that the voltage at the grid connection point UNVP as well as the voltage at the busbar changes accordingly slowly. Here, a period of 400 seconds from the advance timing tv to the reduction time t _{R is} provided. The changes in the voltage shown, in particular the change in the voltage at the busbar Uss shown, are then sufficiently slow that a step transformer can compensate for this voltage change. However, this is achieved with too early a reduction in power, which begins 6.5 minutes too early with the 400 seconds given by way of example and thus leads to a loss. This loss can be illustrated by the indicated energy triangle 630.

In order to avoid this, according to one embodiment, a slow reduction of the reactive power is proposed, as FIG. 7 illustrates.

According to this embodiment, which is illustrated in the diagram of FIG. 7, it is thus proposed that, at the time of reduction t _R, the active power PWEA fed in by the wind park is again reduced stepwise, while the supplied reactive power QWEA is initially kept constant. This leads to a larger inductive cos (cp). The voltage at the network connection point UNVP thereby breaks strongly at the network connection point 304.

In any case, however, the voltage Uss at the busbar 308 breaks only slightly, because it depends only slightly on the active power because of the configuration of the transformer impedance 412 and transformed network impedance 414 and thus experiences only a very small amount due to the stepped reduction of the injected active power PWEA Burglary. The injected reactive power QWEA is then lowered from the reduction time tp according to a ramp or linear function. Both the voltage UNVP at the grid connection point 304 and the voltage Uss at the busbar 308 are thereby gradually increased. The sharp drop in the voltage UNVP at the grid connection point 304 is acceptable and can be increased again by the gradual change of the reactive power QWEA from the reduction time t _R. Their course presents no particular problem for the electrical supply network, in particular the network section 314. However, the increase in the voltage Uss at the busbar 308 is so slow that a step transformer can compensate for this change. This results in only a small effect on the electrical supply network, in particular the voltage at the transformer point 310 can be kept substantially constant.

Basically, here too, the gradual change in reactive power achieves a gradual change in the voltage Uss across the busbar, which can be compensated by a tapped transformer. The solution proposed here, however, creates that nevertheless the active power is only gradually reduced to the reduction time t _R and thus no loss of yield occurs. Here, it was also particularly recognized that the voltage responses at the network connection point 304 on the one hand and a busbar 308 on the other hand are different to active power changes and reactive power changes. Particularly when the network connection point is directly connected to the busbar and the wind park uses or is in cos (cp) control, there are advantages in said method. This has been deliberately taken into account and thus particularly a solution proposed that avoids an abrupt large voltage jump on the busbar and then avoids a corresponding voltage jump on the secondary side 316 of the transformer 312. Because the grid voltage can be considered approximately constant, unwanted voltage differences between the primary and secondary sides of the transformer are avoided.

FIG. 8 shows a variant which is similar to the variant of FIG. Again, the active power PWEA is reduced at the time of reduction tp. However, the reactive power QWEA is also reduced here at the reduction time t _R. As a result, the voltage Uss at the busbar could also increase suddenly, as was the case according to the variant of FIG. Here, however, an additional reactive power control is now performed, especially by an additional device. As a result, additional reactive power can be fed in or provided, which is referred to here as Qz. net becomes. Due to this additional reactive power Qz, this voltage jump of the voltage Uss at the busbar can be counteracted.

In particular, such an additional reactive power control is performed here, which feeds reactive power exactly as much and in a corresponding manner, that the voltage Uss at the busbar does not rise due to the sinking of the active power PWEA and in particular due to the parallel lowering of the reactive power QWEA. This additional reactive power control counteracts so. In this case, the additional reactive power control can provide that the additional reactive power Qz is gradually reduced again after the reduction time t _R , in particular with a decreasing ramp or correspondingly reduced linearly. This then leads to a slow increase in the voltage Uss at the busbar, but this is so slow that a tapped transformer can counteract this. The voltage on the primary side 316 of the transformer 312 can thus be maintained. As a result, the additional reactive power control can then be driven back to a value of 0, so that then no additional reactive power is fed.

The additional reactive power in the reactive power control can basically be provided by the inverters of the wind turbines. The control value can be added to a reactive power value of a phase angle control. The setpoint may be e.g. by a central parking control, which may also be referred to as FCU, or a remote control terminal, which is also referred to as RTU, are predetermined, whereas the phase angle is set by a control or control of a wind turbine.

As further variants, it is proposed that a different wind energy plant or another wind park, a STATCOM or other compensation mechanisms be used as an actuator for generating additional reactive power.

The disadvantage here may be particularly that such an additional reactive power control is complex.

For wind farms that are not directly connected to the busbar, it can be a conflict of interest if they must limit voltage changes both at the grid connection point and at the busbar.

In this case, it can be critical if the wind farm regulates the specification of a cos (cp) reactive power. A regulation of the reactive power can also be used as a control of the reactive power budget of the wind farm can be designated. A particularly critical case is when the reactive power control, especially the given power factor, ie cos (cp), does not fit or does not fit exactly enough to the system. A reactive power budget can also be taken into account if, for example, a network operator specifies a reactive power budget.

In FIG. 9, the voltage at the network connection point (NVP) is temporarily regulated. This indirectly introduces a cos (cp) suitable for the mesh link point, which may also be referred to as a correct cos (cp).

In the exemplary topology according to FIG. 3, this can have a major effect on the voltage at the busbar.

When the park, unlike the exemplary topology of FIG. 3, is directly connected to the busbar, the method proposed according to FIG. 9 leads to particular advantages.

FIG. 9 shows a variant in which a regulation is used which regulates the voltage UNVP at the network connection point 304. For this purpose, it is also assumed here that the injected active power PWEA is suddenly reduced at the reduction time tp. However, a regulation is then provided which initially holds the voltage UNVP at the network connection point. For this purpose, the fed-in reactive power QWEA is reduced correspondingly at the reduction time t _R. The cos (cp) changes as a result. In order to get the cos (cp) back to the previous value, then shortly after the reduction time t _R, the reactive power is slowly ramped up to a setpoint, in particular linear, which corresponds to the initial cos (cp) specification or through which the cos (cp) is realized.

By this measure, the voltage UNVP is initially held at the network connection point, but then also ramped down, so particularly linear down to a later value, the voltage UNVP then retains the network connection point, as soon as the previous cos (cp) was reached again. Of course, this could also be a new setpoint for the cos (cp), if this is desirable for other reasons. At the same time, however, the voltage Uss at the busbar jumps by this measure to the reduction time tp. But she too is going through the gradual linear change in the input reactive power QWEA reduced accordingly.

According to a variant, it is proposed to limit both jumps, ie the jump of the voltage USS at the busbar and the reactive power jump, to a predetermined extent. For this purpose, a small reactive power jump can be allowed in order to subsequently return the reactive power with a ramp. It can thereby be achieved that instead of a large jump in the voltage USS at the busbar, there occurs only a small jump, while at the network connection point also a small voltage jump is accepted. At these two points the tension jumps in different directions, but the jump splits up into two small jumps.

It should be noted, however, that the connection impedance 406 may also be designed completely differently. As an alternative to the topology illustrated in FIG. 3, the network connection point could also be connected to the busbar without or without appreciable impedance, that is, for example, with or without negligibly long connection cable. In that case, the voltage Uss at the busbar would behave like the voltage UNVP at the grid connection point, because the grid connection point and busbar would be electrically equal or at least almost equal. Then the voltage Uss at the busbar would not change abruptly, but only gradually, as shown in Figure 9 for the voltage UNVP at the network connection point. Accordingly, a tapped transformer could counteract this gradual voltage drop.

In that regard, FIG. 9 illustrates, however, that such a regulation for holding the voltage UNVP at the network connection point is not always advisable, but depends on the specific situation, namely also on the transmission behavior from the network connection point to the busbar.

Claims

claims

Method for feeding electrical power into an electrical supply network by means of a decentralized feed unit, in particular by means of at least one wind turbine, wherein

the supply unit is connected to a grid connection point for supplying the electrical power,

the network connection point is connected directly or via a connection connection with a transformer point, for transmitting the electrical power from the network connection point to the transformer point, possibly via the connection connection,

the transformer point is connected to a network section via a transformer, for transferring the electrical power from the transformer point via the transformer to the network section, comprising the steps

Feeding active electrical power into the electrical supply network at the grid connection point,

Feeding electrical reactive power into the electrical supply network at the grid connection point,

Detecting a change to be made in the active power to be injected,

Changing the injected active power according to the detected change to be made, and

Limiting a temporal change of the fed reactive power when changing the injected active power and / or immediately thereafter, to counteract a voltage increase at T ransformatorpunkt and / or in the network section, and / or

temporarily activating a voltage regulation as a function of the change in the injected active power, in order to dynamically regulate the voltage at the reference point or network connection point during and / or immediately after changing the fed active power, at a reference point, in particular at the network connection point, and / or along a trajectory, especially along a ramp.

Method according to claim 1,

characterized in that Limiting the temporal change of the input reactive power is done so that

the fed-in reactive power is changed according to a change function, in particular according to a ramp with a slope limited in magnitude, and / or

a Gradientengre nzwe is given for the fed reactive power, which specifies a magnitude according to the amount of time change of the fed reactive power, in particular so that the input reactive power is changed with a temporal gradient, which does not exceed the magnitude of the Gradientengre nzwe rt.

3. The method according to claim 1 or 2,

characterized in that

the transformer

a primary side, to which the transformer punk is connected, has a secondary side to which the network section is connected, and

the transformer is designed as a step transformer and can set a transmission ratio from the primary side to the secondary side, thereby controlling a voltage level at the T ransformatorpunkt and / or at the network section, wherein

the limitation of the change over time of the supplied reactive power is such that a voltage change resulting from the change in the fed-in reactive power at the transformer point or at the primary side is so slow that the tapped transformer can compensate a resulting voltage change.

4. The method according to any one of the preceding claims,

characterized in that

for the fed active power no limitation of a temporal change is provided. 5. The method according to any one of the preceding claims,

characterized in that

the temporarily activated voltage regulation is active only when changing the active power fed in and / or immediately thereafter, and the voltage at the network connection point is wholly or partly reduced to a value which is changed by changing the fed active power the voltage at the network connection point immediately before changing the fed active power, wherein the temporarily activated voltage regulation is carried out in particular when the network connection point is directly connected to a transformer point.

Method according to one of the preceding claims,

characterized in that

for the supply of electric reactive power, a power factor control (cos (p) control) is provided, in which the reactive power fed in response to the injected active power is adjusted so that a predetermined power factor (cos (cp)) results,

when changing the injected active power

the fed-in reactive power is initially changed at the same time so that the power factor remains unchanged,

a new value for the power factor is then given as a function of a voltage which changes as a result of the change in the active power and reactive power, or the expected voltage change at the network connection point, and

the power factor is returned to the predetermined power factor, the feedback being delayed and / or performed over a time function.

Method according to one of the preceding claims,

characterized in that

for the supply of electric reactive power, a power factor control (cos (cp) control) is provided, in which the reactive power fed in response to the injected active power is adjusted so that a predetermined power factor (cos (cp)) results,

when changing the injected active power it is changed along a ramp or trajectory,

when changing the applied active power along the ramp or trajectory, a voltage regulation is used to keep the voltage at the reference point constant, at least to counteract a voltage change by changing the applied active power, the voltage regulation for this purpose sets the reactive power and the power factor control (cos ( cp) control) is deactivated during this voltage regulation, after the change in the injected active power, a reactive power value resulting from the voltage regulation is fed to a new reactive power value, which would result from the deactivated power factor control (cos (cp) control)

- The reactive power value is guided, in particular via a ramp and / or trajectory to the new reactive power value, and

the deactivated power factor control (cos (cp) control) is activated as soon as the reactive power value has reached the new reactive power value.

8. The method according to any one of the preceding claims,

characterized in that

changing the injected active power is performed via a real power jump, whereas

the changing of the reactive power over a time function, in particular via a ramp function is performed. 9. The method according to any one of the preceding claims,

characterized in that the limitation of the temporal change of the input reactive power of the type and / or size of the connection connection is dependent.

10. Wind turbine for feeding electrical power into an electrical supply network wherein

the network connection point is connected directly or via a connection connection with a transformer point, for transmitting the electrical power from the network connection point via the connection connection to the transformer point,

the transformer point is connected via a transformer to a network section, for transmitting the electrical power from the transformer point via the transformer to the network section, and the wind turbine comprises

an inverter controller for controlling a supply of reactive electric power to the electric utility network at the grid connection point, an input interface for detecting a change to be made in the active power to be fed, wherein

the inverter controller is prepared to control or limit a time change of the injected reactive power when changing the injected active power and / or immediately thereafter to counteract a voltage increase at T ransformatorpunkt and / or in the network section and / or

the inverter control is prepared to carry out a temporary activation of a voltage regulation as a function of changing the injected active power in order to carry out a voltage regulation at a reference point, in particular at the network connection point, during and / or immediately after changing the fed active power To dynamically balance and / or guide along a trajectory, in particular along a ramp.

1 1 wind turbine according to claim 10,

characterized in that

the wind power plant, in particular the inverter control, is prepared to perform a method according to one of claims 1 to 9.

12. Wind park comprising several wind turbines, wherein

Wind turbine according to claim 9 or 10 are used, and / or - the wind park is prepared to perform a method according to one of claims 1 to 9.

13. A feed arrangement for feeding electrical power into an electrical supply network, comprising

a decentralized feed-in unit, in particular a wind turbine, a grid connection point, and

a transformer with a transformer point, where

the network connection point is connected directly or via a connection connection with the transformer point, for transmitting the electrical Power from the network connection point via the connection to the transformer point,

the transformer point is connected to a network section via the transformer, for transferring the electrical power from the transformer point via the transformer to the network section, and comprising the feed arrangement

an inverter controller for controlling a supply of reactive electric power to the electric utility network at the grid connection point,

an input interface for detecting a change to be made in the active power to be fed, wherein

the inverter control is prepared for limiting a time change of the supplied reactive power when changing the fed active power and / or immediately thereafter to counteract a voltage increase at T ransformatorpunkt and / or in the network section, and / or

the inverter control is prepared to perform a temporary activation of a voltage regulation as a function of changing the injected active power in order to carry out a voltage regulation at a reference point, in particular at the network connection point, during and / or immediately after changing the injected active power, by the voltage at the reference point or network connection point dynamically correct, and / or to guide along a trajectory, in particular along a ramp.

14. Feed arrangement according to claim 13,

characterized in that the feed arrangement, in particular the inverter control,

is prepared to carry out a method according to one of claims 1 to 9,

as feed unit comprises a wind turbine according to claim 10 or 11, and / or

as feed unit a wind park according to claim 12.