Patent Application
20200176993
The present invention relates to a method for operating a wind farm. The present invention also relates to a corresponding wind farm.
Wind farms are known in principle and comprise a number of wind power installations. The wind power installations of the wind farm then together feed electric power into an electrical supply grid, in particular by way of a common grid connection point.
With the increasing dominance of decentralized feeding-in units such as wind power installations or wind farms in the electrical supply grid, it is becoming increasingly important that wind farms also make a contribution to controlling the electrical supply grid. Such methods are also already known, and for example, on a frequency-dependent basis, wind farms feed in a correspondingly greater or lesser amount of reactive power. Reducing the amount of power fed in or briefly increasing the real power fed in also comes into consideration as a way in which wind farms control or support the electrical supply grid.
However, a greater proportion of decentralized, in particular converter-controlled, feeding-in suppliers also has the effect of changing the way in which the electrical supply grid behaves. Known supporting measures, such as changing the power on a frequency-dependent basis can then sometimes no longer be suitable, because the changing described possibly makes the electrical supply grid react fundamentally differently.
The reaction to a changed grid behavior could be in principle a switchover between different controllers. The laid-open patent application DE 10 2013 207 264 A1 proposes a controller switchover for this. It is questionable, however, whether with such a controller switchover all grid changes or even changed grid requirements can be addressed.
The German Patent and Trademark Office has also searched the following prior art in the priority application relating to the present application: DE 10 2006 050 077 A1, DE 10 2014 214 151 A1 and the Annual Report 2016 of the Technical University of Braunschweig, Institute for High Voltage Technology and Electrical Power Systems (elenia), pages 54-56.
Provided is a method intended for operating a wind power installation or a wind farm comprising a number of wind power installations for exchanging electrical power between the wind farm and an electrical supply grid. In particular, it is concerned with controlling the feeding into the electrical supply grid by the wind farm, but it also comes into consideration that the wind power installation or the wind farm takes power from the electrical supply grid, particularly in the case of special grid supporting operations. To this extent, the method relates generally to the operation of a wind power installation or a wind farm for exchanging electrical power between the wind power installation or the wind farm and the electrical supply grid.
The method assumes that each of the wind power installations comprises one or more feeding-in devices. Particularly, at least one converter or inverter may be provided for each wind power installation. Preferably, to realize higher power, a number of converters or inverters are connected in parallel. In that case, a wind power installation then has a number of feeding-in devices. If the method only relates to the operation of one wind power installation, the latter has a number of feeding-in devices, but for the operation of a wind farm it is preferably the case that a number of feeding-in devices are provided in each wind power installation. A distinction between converter and inverter is only of secondary importance here and, unless anything else is explained or anything else is obvious, any explanations concerning converters should also be understood as explanations concerning inverters, and vice versa. As long as they do not concern aspects peculiar to a wind farm, any explanations of a wind farm that are described hereafter should also be applied analogously to an individual wind power installation.
It is also assumed that the wind power installation or the wind farm is connected to the electrical supply grid by way of a grid connection point and the power is exchanged by way of the grid connection point. The wind power installations of the wind farm therefore share a common grid connection point. In principle, the method is however also suitable for controlling a number of wind farms, each wind farm then being able to have a grid connection point of its own.
It is proposed that one or more of the feeding-in devices operate as voltage-impressing units. They consequently exchange power with the electrical supply grid in a voltage-impressing manner, in particular they feed into the electrical supply grid in a voltage-impressing manner. It is also proposed that one or more of the feeding-in devices operate as current-impressing units. They consequently exchange power with the electrical supply grid in a current-impressing manner, in particular they feed into the electrical supply grid in a current-impressing manner.
It is therefore proposed that the wind farm comprises one or more voltage-impressing units and one or more current-impressing units for feeding in. Such feeding-in units, that is to say voltage-impressing on the one hand and current-impressing on the other hand, may in principle operate differently. Operating as a voltage-impressing unit means particularly impressing a preset voltage. In simple terms, a voltage is fed back there, and to this extent forms the controlled variable, and the voltage-impressing unit correspondingly attempts to correct to this voltage, therefore to set this voltage.
By contrast, operating as a current-impressing unit means controlling a current in a closed-loop manner. In this case, therefore, a current is particularly output and measured and this measured value is compared as an actual value with a setpoint value and, depending on the difference between the setpoint value and the actual value, is correspondingly controlled in a closed-loop manner.
Consequently, some units operate in a voltage-impressing manner and others in a current-impressing manner. Preferably, it is proposed that the same units are the same with regard to their hardware, but they can optionally operate in a current-impressing or voltage-impressing manner, particularly by corresponding activation. However, it also comes into consideration that units which are voltage-impressing or current-impressing from the outset are provided, and they therefore differ in their hardware or their construction but can be activated or used according to requirements. In this case, it is proposed here that both voltage-impressing units and current-impressing units are activated in order to control the exchange of the power with the electrical supply grid.
With such combined operation, in which some units operate in a voltage-impressing manner and others in a current-impressing manner, it is also possible in principle to make allowance for changed requirements for the electrical supply grid. Particularly an electrical supply grid which, for example because of its topology, is itself poorly able to maintain its own voltage can be supported by means of voltage-impressing feeding in or voltage-impressing power exchange.
The voltage-impressing operation of some units allows a corresponding feed-in current, and consequently power, to be fed into the electrical supply grid particularly stably. This may, however, also involve feed-in current or power being taken from the grid, particularly for a short period of time.
In principle, however, voltage-impressing operation and current-impressing operation also allow the sharing of special closed-loop control tasks for which voltage-impressing operation and current-impressing operation are suited differently well. Examples of this are also given below.
It is particularly provided that the voltage-impressing units and the current-impressing units also operate in a voltage-impressing and current-impressing manner during undisturbed operation of the electrical supply grid. Undisturbed operation may also be referred to as normal operation. Undisturbed operation is consequently operation without disturbances, the term disturbances referring to significant disturbances, such as an interruption in the grid, a short-circuit in the electrical supply grid and in particular a grid recovery or black start of the electrical supply grid or a portion of the grid. The use of voltage-impressing and current-impressing units in a wind farm is consequently proposed generally, and is not intended to be restricted to a specific mode; it is particularly not intended to be restricted to a specific black-start mode, but is expressly intended for a normal operating mode. Undisturbed operation or normal operation is to this extent an operating mode in which the wind power installation or the wind farm is in operation to generate power from the wind and feed it into the electrical supply grid at the level at which the power can be generated from the wind. Such normal operation may be interrupted by special supporting modes of operation for which it is also advisable to combine the voltage-impressing and current-impressing units. The voltage- and current-impressing units are however not only switched on for special operating modes, but are already active in a voltage-impressing and current-impressing manner during normal operation.
It is preferably proposed that, for operation of the wind power installation or the wind farm, voltage-impressing units exchange power with the electrical supply grid in a voltage-impressing manner and at the same time current-impressing units exchange power with the electrical supply grid in a current-impressing manner. Exchanging power with the electrical supply grid in a voltage-impressing and current-impressing manner are consequently intended to take place at the same time. Therefore, it is not a case of choosing one or other of the two modes, but rather that they complement one another. Depending on the application, in particular depending on the situation in the electrical supply grid, it also comes into consideration that the ratio of operating in a voltage-impressing manner and operating in a current-impressing manner can be varied.
In any event, the current-impressing units in current-impressing operation respectively control a feed-in current that is to be fed into the electrical supply grid. The term feed-in current is also used here when electrical power is taken from the electrical supply grid. For this case, the feed-in current may be negative or its phase angle may be chosen in such a way that power is taken from the electrical supply grid.
In voltage-impressing operation, the voltage-impressing units control a feed-in voltage for feeding into the electrical supply grid. Here, the feed-in voltage is therefore the decisive target variable or controlled variable. In current-impressing operation, closed-loop control is therefore directed at the feed-in current and in voltage-impressing operation it is directed at the feed-in voltage. Current-impressing units are those that operate in a current-impressing manner and voltage-impressing units are those that operate in a voltage-impressing manner. However, this characteristic does not necessarily have to be permanently assigned to the respective unit, but may possibly also be achieved by corresponding changing of the control.
Preferably, the current-impressing units operate in such a way that they adapt themselves in current-impressing operation to the feed-in voltage, which particularly in a wind farm is present at the grid connection point. They therefore control the feed-in current, and the feed-in voltage forms a boundary condition for this, for which allowance is made. It is not attempted however to correct specifically to a voltage value, at least not in the sense that a voltage setpoint value is preset. This however does not affect the possibility of a higher-level closed-loop control responding to a voltage or voltage deviation, such as for example, to set a phase angle of the current in dependence on the voltage. In the case of this example, although the setting of the phase angle also has the purpose of having an effect on the level of the voltage, the closed-loop control or type of feeding in that is carried out here by the current-impressing unit is to correct to such a setpoint current.
In voltage-impressing operation, the voltage-impressing units respectively feed in a current that is required for controlling the feed-in voltage or is obtained by controlling the feed-in voltage. For voltage-impressing operation, consequently a setpoint current is not provided for the corresponding voltage-impressing units, but a setpoint voltage. It is attempted to achieve this voltage by using closed-loop control techniques, and that will result in a feed-in current. If, therefore, the feed-in voltage for example falls below a corresponding setpoint value for it, or otherwise deviates from a setpoint value, the respective voltage-impressing unit in voltage-impressing operation attempts to counteract this fall, and that may lead to a correspondingly increased feed-in current. Here, too, such a feed-in current may be of the kind that electrical power is taken from the grid.
According to one embodiment, the method operates in such a way that, in the case of an overfrequency in the electrical supply grid, the power fed into the electrical supply grid by the wind farm is reduced. For this purpose, the method may operate in such a way that the voltage-impressing units first reduce their feed-in power in an initial time period, in order thereby to bring about the reduction in the power fed in. With regard to the current-impressing units, the method may operate in such a way that they follow the voltage-impressing units by reducing their feed-in power, in order to bring about the reduction in the power fed in in a subsequent time period following the initial time period, when the reduction of the power fed in at the wind farm has assumed a steady value. Then, the voltage-impressing units are operated to carry out a process of maintaining the voltage and, in particular in the subsequent time period, leave the reduction of the power fed in by the wind farm to the current-impressing units. As a result, the voltage-impressing units undertake the acute first open-loop control or first closed-loop control by their voltage-impressing operation.
A frequency change in the electrical supply grid means initially that the frequency of the voltage in the electrical supply grid changes. Therefore, first an effect on the voltage is detected. This effect on the voltage by this frequency change consequently makes itself noticeable directly at the voltage-impressing units, since they record the voltage and attempt to maintain its setpoint value. It is not, or not only, rms voltages that are concerned here, but particularly also instantaneous values. If the frequency of the actual voltage changes in comparison with the frequency of the setpoint voltage, the phase angle between the actual voltage and the setpoint voltage also changes. That leads directly to a voltage deviation in each instantaneous value. The voltage-impressing unit, which is operating in voltage-impressing mode, can react to this immediately, before a higher-level closed-loop control determines a changed setpoint value, dependent on a frequency change.
In rather more simple terms, a voltage-impressing unit reacts to a frequency change in a way similar to a directly connected synchronous generator. In any event, a voltage-impressing unit behaves more similarly to a synchronous generator than a current-impressing unit.
The current-impressing units may undertake the power reduction somewhat later, in particular because they receive a reduced current setpoint value. As soon as the current-impressing units have undertaken this power reduction, this may also have an effect on the frequency and voltage, and the controlling component of the voltage-impressing units may then be reduced again. In an ideal case, the power reduction by the voltage-impressing units may as a result be reduced to zero. However, it can often happen that such operation at an overfrequency only occurs very briefly, and the behavior of the voltage-impressing units to this extent remains relevant. In other words, here the voltage-impressing units essentially perform the dynamic closed-loop control, whereas the current-impressing units tend rather to assume the steady state.
According to one embodiment, it is proposed that, in the case of overfrequency in the electrical supply grid, power is taken by the wind farm from the electrical supply grid and in particular is consumed. This embodiment consequently goes one step further than merely reducing power in the case of an overfrequency, in that it is additionally proposed even to take power from the supply grid. The power may then be consumed for example in a chopper system.
It is in this case proposed that the voltage-impressing units are operated to take up the power from the electrical supply grid, at least in the initial time period. Consequently, it is proposed even for the case where power is taken from the grid that at least initially the voltage-impressing units essentially undertake this task here. In principle, it is also provided here that the voltage-impressing units undertake the closed-loop control, particularly at the beginning and consequently at the transition to such an overfrequency situation. Since, however, such a case in which electrical power should even be taken from the grid often only exists for a short transitional time period, this closed-loop control will then also essentially be undertaken overall by the voltage-impressing units.
It has consequently been recognized that, even for the case of power take-up, the voltage-impressing units are particularly well suited for this task.
It is preferably provided that the wind farm feeds into the electrical supply grid with a feed-in voltage. For this purpose, a check is made for phase jumps in the electrical supply grid. Such a phase jump of the grid voltage leads to a phase shift in the grid voltage, that is to say in a temporal respect. The absolute phase position of the grid voltage therefore changes abruptly, so that for example a zero transition of the grid voltage is shifted by a millisecond in comparison with the previous rhythm.
Such a phenomenon can be attributed particularly to the behavior of one or more synchronous generators that are coupled directly to the electrical supply grid. Such synchronous generators have an angular displacement, which depends on the stator current and determines the position of the synchronous generated voltage. A change of the stator current can consequently cause a change of the rotor, and consequently of the phase position of the synchronous generated voltage.
In the case of this embodiment, it is thus proposed that a check is made for corresponding phase jumps. In this case, the method operates in such a way that, in the case of a phase jump, the voltage-impressing units in a first step initially feed in real and/or reactive power, in order to maintain the frequency and/or phase position of the feed-in voltage. The voltage-impressing units therefore react immediately to such a phase jump and in actual fact attempt to resist it. As a result, the phase jump can initially be somewhat counteracted. As a result, an excessive phase jump can also possibly be prevented, and consequently the possible loss of synchronism of a corresponding synchronous generator coupled directly to the electrical supply grid can be prevented. Particularly, this behavior of the voltage-impressing units may at least somewhat stabilize the synchronous generator that has caused the phase jump, though it is also possible that a number of synchronous generators may have caused the phase jump. For this behavior, it is also possible to do without an express check for a phase jump, because an allowance for it is already achieved by the voltage-impressing behavior. For further steps, and consequently the proposed concept overall, a check for a phase jump is however advisable.
If a phase jump is then detected, in a second step a reference frequency, and also or alternatively a reference voltage, of the changed grid voltage may be controlled as a dependent variable. Preferably, only a reference voltage is preset, including in this case the corresponding reference frequency, that is to say in its specific signal form. By specifying this reference frequency or reference voltage in this way, it is thus possible specifically to preset a corresponding reference value that is advantageous for the stability of the electrical supply grid. It comes into consideration in particular that the reference voltage is preset in such a way that the transition from the previous voltage profile before the occurrence of the phase jump to the voltage profile after the occurrence of the phase jump takes place gradually, so that an overreaction in the electrical supply grid is prevented. In this case it also comes into consideration that the behavior of the grid itself is observed for this phase jump, and is used for presetting the reference voltage. For example, the electrical supply grid may react to the phase jump of its own accord, that is to say by feeding-in suppliers or grid participants other than the wind farm reacting to the phase jump, and for example change the phase back or change it in some other way or change a frequency. Such an effect can be observed, in particular measured, and the presetting of the reference frequency or reference voltage preferably takes place at the wind farm in such a way that particularly a resonance in the electrical supply grid, caused by other countermeasures, is counteracted, and therefore such a resonance is prevented.
Finally, it is then provided that the voltage-impressing units control the feed-in voltage as a dependent variable in a way corresponding to the reference frequency or reference voltage. The reference frequency or reference voltage consequently form setpoint values for the closed-loop control of the voltage-impressing units. In particular, a reference frequency or reference voltage may also be preset in such a way that this can be carried out well for the implementation by the voltage-impressing units. By presetting the reference frequency or reference voltage in such a specifically deliberate way, it can therefore be prevented that reactive and/or real currents resulting in implementation do not become too great.
According to one embodiment, it is proposed that the wind farm realizes a voltage-impressing mode with power feed-in by a combined control of the voltage-impressing units and the current-impressing units. It is proposed for this purpose that the voltage-impressing units essentially perform the voltage impression and also or alternatively dynamic compensating operations, and that the current-impressing units essentially perform a steady-state power feed-in.
The voltage-impressing units react particularly quickly to voltage changes or voltage deviations, and are consequently particularly quick closed-loop control tools.
This is utilized in this mode of combined control. Consequently, the voltage-impressing units can provide the voltage impression, but also initiate quick closed-loop control operations.
The current-impressing units, which essentially perform closed-loop control current-dependently, react rather more indirectly to changes in the electrical supply grid in comparison with the voltage-impressing units, in particular indirectly to voltage changes. That is to say that the current-impressing units react essentially to current changes, which may be a consequence of voltage changes, and they react to changed setpoint values, which may for example be preset for them by a higher-level control device. It can in this way be ensured however that the current-impressing units can feed in power comparatively stably and steadily.
According to one embodiment, each feeding-in device is characterized by a power coefficient, in particular a rated power. It is proposed for this purpose that, with reference to the respective power coefficients, the wind farm has a greater proportion of current-impressing units than voltage-impressing units. Therefore, the current-impressing units dominate in the wind farm. Consequently, a good, stable power feed-can be achieved, whereas at the same time a high level of control dynamics can be ensured by the voltage-impressing units.
Preferably, the voltage-impressing units form a proportion of at least 2% and at most 25%, preferably of at most 15% and in particular of at most 10%. These proportions relate to a 100% proportion, which denotes the sum of the power coefficients of all of the feeding-in devices. It is therefore proposed that less than one third, on the basis of the power coefficients, of voltage-impressing units are in the wind farm. Just a proportion of 15% or 20% can ensure the stated concept, that is to say of achieving a good, stable power feed-in by as many current-impressing units as possible, but nevertheless still maintaining a sufficient proportion of voltage-impressing units for purposes of quick closed-loop control.
Preferably, the voltage-impressing units have at least one first droop and the current-impressing units have at least one second droop. Each droop respectively describes a relationship, particularly a linear relationship, between an electrical voltage of the wind farm and a reactive power that is to be fed in or has been fed in. Each droop may also alternatively describe a relationship between a frequency of the wind farm and a real power that is to be fed in or has been fed in. It is proposed for this purpose that the first droop has a smaller gradient than the second droop. In simple terms, the droop of the current-impressing units is steeper than the droop of the voltage-impressing units.
The first droop, that is to say the droop of the voltage-impressing units, particularly describes a relationship between the electrical voltage of the wind farm respectively at the output of the voltage-impressing unit and the reactive power fed in by this voltage-impressing unit. The voltage-impressing unit then sets the electrical voltage particularly at its output, which is consequently a voltage of the wind farm, in dependence on the reactive power that it has itself fed in. In this case, a setpoint value for this voltage is particularly set, that is to say in a way corresponding to the first droop in dependence on the reactive power fed in.
The second droop, that is to say the droop of the current-impressing units, in this case relates to a dependence of the reactive power to be fed in on the electrical voltage of the wind farm. Here, too, the electrical voltage at the output of the current-impressing unit concerned comes into consideration, but here thus forms the input variable. Dependent on this voltage, and correspondingly the second droop, the reactive power to be fed in is then determined, and is preset particularly as a reactive-power setpoint value. Each current-impressing unit then feeds as far as possible a reactive power according to this reactive-power setpoint value.
For the case where the droop presets a dependence between frequency and real power, a frequency setpoint value is preset for the voltage-impressing units in a way corresponding to the first droop in dependence on the real power fed in. For the current-impressing units, a setpoint value for the real power is preset according to the second droop on a frequency-dependent basis.
These different gradients of the droops achieve the effect that the voltage-impressing units initially carry out closed-loop control quickly and with great manipulated variables, whereas the current-impressing units to this extent have a weaker gain, and as a result leave the dynamic closed-loop control essentially to the voltage-impressing units.
According to one embodiment, it is proposed that the voltage-impressing units respectively have at least one first reactive-power droop, which respectively describes a relationship between an electrical voltage to be provided by the voltage-impressing unit in the wind farm and a reactive power that is fed in by the voltage-impressing unit. The first reactive-power droop therefore indicates at what level the voltage-impressing unit is in each case to provide an electrical voltage in the wind farm in dependence on its fed-in reactive power.
It is also proposed that the current-impressing units respectively have at least one second reactive-power droop, which respectively describes a relationship between an electrical voltage recorded in the wind farm and a reactive power to be fed in by the current-impressing unit. The second reactive-power droop therefore indicates how much reactive power the current-impressing unit is to feed in, in dependence on the recorded electrical voltage.
It is thus proposed for this purpose that the first reactive-power droop, that is to say that of the voltage-impressing units, has a smaller gradient than the second reactive-power droop, that is to say that of the current-impressing units. The first droop in this case indicates however how much voltage is to be set in dependence on the reactive power fed in. Just a small value of reactive power fed in can consequently lead to a high voltage value to be set. Therefore, for this the voltage-impressing unit or its closed-loop control is more dominant than the current-impressing units.
Also or alternatively, according to one embodiment, it is proposed that the voltage-impressing units respectively have at least one first real-power droop. The real-power droop respectively gives a relationship between a frequency to be provided in the wind farm by the voltage-impressing unit and a real power that is fed in by the voltage-impressing unit. The first real-power droop therefore indicates how high the frequency should be chosen, in dependence on the real power fed in.
For this purpose, the current-impressing units respectively have at least one second real-power droop, which respectively describes a relationship between a frequency recorded in the wind farm and a real power to be fed in by the current-impressing unit. The second real-power droop therefore indicates at what level the current-impressing unit is intended to feed in amounts of real power in dependence on the recorded frequency.
It is thus proposed for this purpose that the first real-power droop has a smaller gradient than the second real-power droop. Here, too, the frequency is therefore set at the voltage-impressing units in dependence on the real power and, due to the small gradient of the first real-power droop, correspondingly strong frequency deviations are obtained when there are small real-power deviations. Consequently, here, too, the voltage-impressing unit is in each case a quick, dynamic closed-loop control. The voltage-impressing units are to this extent also intended here for the particularly dynamic closed-loop control operations.
It is preferably also proposed that the voltage-impressing units are controlled by the first reactive-power droop in such a way that it controls an electrical voltage to be provided by the voltage-impressing unit in the generating grid in dependence on a reactive power fed in by the voltage-impressing unit in such a way that the reactive power fed in is controlled to a value of almost zero. At least, a steady system deviation that is as small as possible is intended to be achieved. Preferably, though the voltage-impressing units are quickly controlling units, in return it is intended to achieve as far as possible closed-loop control to a value of almost zero in the steady state, at least for the reactive power. A small system deviation, for example in the range of 5% with reference to the respective rated power, may be advisable and create a degree of latitude for a closed-loop control by the current-impressing units.
A wind farm is also proposed. Such a wind farm is intended for exchanging electrical power with an electrical supply grid by way of a grid connection point and comprises a number of wind power installations. Each of the wind power installations has one or more feeding-in devices. A feeding-in device may be a converter or an inverter or an arrangement of converters or inverters.
One or more of the feeding-in devices operate as voltage-impressing units and exchange power with the electrical supply grid in a voltage-impressing manner. One or more of the feeding-in devices operate as current-impressing units and exchange power with the electrical supply grid in a current-impressing manner. The wind farm therefore has voltage-impressing and current-impressing units and this enables it to achieve the advantages that have already been described above in connection with the method for operating a wind farm.
Preferably, each feeding-in device is characterized by a power coefficient, which may particularly be the rated power of the feed-in device, and it is provided that, with reference to the respective power coefficients, the wind farm has a greater proportion of current-impressing units than voltage-impressing units. Preferably, the wind farm has at least 2% and at most 25%, preferably at most 15% and in particular at most 10%, voltage-impressing units. The values relate in this case to a 100% proportion, which denotes the sum of the power coefficients of all of the feeding-in devices of the mind farm. It is therefore proposed that the current-impressing units dominate.
According to a further embodiment, it is proposed that each wind power installation or each feeding-in device has at least one control device and at least one droop is respectively stored in the control device. It is proposed for this purpose that the voltage-impressing units have at least one first droop and the current-impressing units have at least one second droop. Each droop in this case respectively describes a relationship, in particular a linear relationship, between an electrical voltage of the wind farm and a reactive power that is to be fed in or has been fed in. Another droop respectively describes a relationship, particularly a linear relationship, between a frequency of the wind farm and a real power that is to be fed in or has been fed in. It is proposed overall for this purpose that the first droop has a smaller gradient than the second droop. As a result, the control dominances can be split between voltage-impressing units and current-impressing units. In particular, a first droop with a smaller gradient than the second droop has the effect that the voltage-impressing units dominate in dynamic closed-loop control operations.
Preferably, the second droop has at least twice the gradient in comparison with the first droop, in particular at least three times and preferably at least four times the gradient in comparison with the first droop. Because the second droop is at least twice as steep, preferably at least three times as steep and in particular at least four times as steep, a voltage-impressing unit can dominate significantly over a current-impressing unit with regard to their closed-loop control characteristics.
According to a further embodiment, it is proposed that each wind power installation or each feeding device has at least one control device and at least one droop is respectively stored in the control device, the voltage-impressing units respectively having a first droop and the current-impressing units having at least one second droop, each droop respectively describing a relationship, particularly a linear relationship, between an electrical voltage of the wind farm and a reactive power that is to be fed in or has been fed in, or that each droop describes a relationship between a frequency of the wind farm and a real power that is to be fed in or has been fed in, and it being the case for both types of droop that the first droop has a smaller gradient than the second droop. Consequently, the voltage-impressing units can provide a quicker dynamic closed-loop control.
Preferably, in at least one of the wind power installations, in particular in all of the wind power installations of the wind farm, there is respectively provided at least one voltage-impressing unit and one current-impressing unit. As a result, each wind power installation can itself provide benefits of combining current-impressing and voltage-impressing units. Then there is also the fact that, as a result of the current-impressing units, power can particularly be fed in stably and steadily, and consequently the provision of current-impressing units in each wind power installation can also achieve the effect that each wind power installation can therefore feed in well the power generated by it. At the same time, however, each wind power installation is also prepared to achieve dynamic closed-loop control by at least one voltage-impressing unit.
Furthermore, a wind farm can also be planned in an easy way, by in principle each wind power installation having an advantageous proportion of current-impressing and voltage-impressing units. If these proportions are chosen to be the same for each wind power installation, the distribution in each wind power installation at the same time corresponds to the distribution in the wind farm. This also has the advantage that, whenever one wind power installation fails, the distribution of current- and voltage-impressing units in the wind farm remains unchanged. Adding further wind power installations to the wind farm also does not cause the split between voltage-impressing units and current-impressing units to change, even if such an added wind power installation has current- and voltage-impressing units, in particular in the ratio as is provided in the wind farm.
It also comes into consideration, however, that not every wind power installation has current- and voltage-impressing units, but only some of them, and other wind power installations have only current-impressing units. As a result, it is possible to allow for the situation where preferably only a very small proportion of voltage-impressing units are provided in the wind farm. If, for example, only 5% of voltage-impressing units are provided, it could be the case for example in a wind farm with wind power installations which respectively have 10 feeding-in devices that alternately one wind power installation has only current-impressing units and another respectively has one voltage-impressing unit and nine current-impressing units.
Preferably, the wind farm is prepared for the purpose of performing a method according to at least one of the embodiments described above of a method for operating a wind farm. The wind farm is particularly prepared for the purpose that it has corresponding control units in the wind power installations and/or for each feeding-in device. The corresponding method steps may be implemented in these control devices.
It is preferably provided that the wind farm has a central control unit, in order to coordinate the wind power installations and in particular the method for operating the wind farm. The coordination may advantageously take the form that the central control unit provides setpoint values, in particular frequency and voltage setpoint values, for the voltage-impressing units and provides real-power and reactive-power setpoint values to the current-impressing units. By providing such setpoint values, in principle a coordination of the wind farm can be achieved, without particularly high requirements for short reaction times and reacting quickly to grid states having to be ensured by this central control unit. It can achieve the coordination of the wind farm, but very quick closed-loop control processes can be achieved by each wind power installation itself.
Preferably, the wind farm has at least one wind power installation according to one of the following embodiments.
Also proposed is a wind power installation for exchanging electrical power with an electrical supply grid, by way of a grid connection point, comprising:
a number of feeding-in devices, wherein
one or more of the feeding-in devices operate as voltage-impressing units; and
one or more of the feeding-in devices operate as current-impressing units, wherein
the voltage-impressing units and the current-impressing units also operate in a voltage-impressing and current-impressing manner during undisturbed operation of the electrical supply grid.
The advantages are evident from the above explanations of the wind farm and/or the method for operating the wind power installation. Reference is also made to the advantages of the following embodiments of a wind power installation.
Preferably, the wind power installation is characterized in that,
each feeding-in device is characterized by a power coefficient, in particular a rated power, and in that, with reference to the respective power coefficients,
the wind power installation has a greater proportion of current-impressing units than voltage-impressing units, wherein preferably,
with reference to a sum of the power coefficients of all of the feeding-in devices as a 100% proportion, the voltage-impressing units have a proportion of at least 2% and at most 25%, preferably a proportion of at most 15% and in particular a proportion of at most 10%.
According to one embodiment, a wind power installation is proposed, characterized in that,
each feeding-in device has at least one control device and at least one droop is respectively stored in the control device, wherein
the voltage-impressing units have at least one first droop, and
the current-impressing units have at least one second droop, wherein
each droop respectively describes a relationship, particularly a linear relationship,
between an electrical voltage of the wind power installation and a reactive power that is to be fed in or has been fed in, or
between a frequency of the wind power installation and a real power that is to be fed in or has been fed in, and wherein
the first droop has a smaller gradient than the second droop.
Preferably, the wind power installation is configured for the purpose of performing a method according to one of the embodiments above of the operating method, and/or of being used in a wind farm according to an embodiment above.
The invention is explained in more detail by way of example below on the basis of exemplary embodiments with reference to the accompanying figures.
The three other power cabinets 502-504 are provided as current-impressing power cabinets, that is to say as current-impressing units. Each current-impressing power cabinet may also comprise a number of current-impressing units. Each of these power cabinets is intended to be able to feed in a greater reactive power Q than the first power cabinet 501. By way of example, a reactive power value of in each case 0.3 MVA is indicated here as a dimensioning variable for each of the power cabinets 502-504. These four power cabinets 501-504 may together feed a reactive power Q amounting to 1 MVA into the grid 560 shown by way of example. This split also serves primarily for purposes of illustration.
Wherein the voltage-impressing power cabinet 501 adapts the voltage deviation dU by way of a setpoint voltage in accordance with the reactive power Q that is obtained and the current-impressing switchgear cabinets adapt the reactive power Q in accordance with the recorded voltage deviation dU. This voltage deviation can be recorded at output terminals of the inverter.
In this case, the graph of
Analogously, here, too, the voltage-impressing power cabinet 501 feeds in a voltage at the frequency f in accordance with the real power obtained and the current-impressing switchgear cabinets feed in the real power in accordance with the recorded frequency or frequency deviation.
The control device 706 shown has a rectifying unit 708 with a connected DC link 710 and a downstream inverter 712. The rectifying unit 708 with the DC link 710 and the inverter 712 can also be referred to collectively as a converter.
The rectifying unit 708 is fed by a generator of the wind power installation, which is only indicated in
The central farm control unit 702 can transmit a power setpoint value Psetpoint to each wind power installation 704. If the wind power installations 704 are each of the same size, there is also the possibility of these values being the same. In actual fact, these power setpoint values Psetpoint are only meant to be representative of possible power setpoint values, which can even have different values, or to be able to be transmitted as relative values, for example percentage values.
In the depicted control device 706 of one wind power installation 704, it is illustrated that this power setpoint value Psetpoint is transferred to the inverter controller 720. There is however also the possibility of other control or evaluation architectures in the wind power installation 704.
The control device 706 shown in
It can also operate in a current-impressing manner, by essentially orienting itself to the recorded and fed-back current i(t) and generating a corresponding current signal, that is to say operating such that its output is particularly controlled as a dependent variable to the current according to a current setpoint value.
This current and voltage impression particularly involves the specific sinusoidal signal that the inverter generates or attempts to generate being preset. Correspondingly, the instantaneous values u(t) and i(t) for the voltage and current for feedback are shown in
The inverter 712 can therefore operate in different ways and in so doing generate a power and feed it into the grid 724 by way of a farm transformer 722. The other wind power installations can use the same farm transformer 722 to feed into the electrical supply grid 724.
Furthermore, a respective isolating switch 726 and 728 is provided both for the farm and with respect to the electrical supply grid 724. Indicated upstream of the isolating switch 726 is a connecting line to the other wind power installations 704, this representing a farm grid 730 here.
Shown for illustrative purposes are a voltage-impressing unit 808 and a current-impressing unit 810, which themselves are also part of the connecting structure 800.
The voltage-impressing unit 808 has a voltage-impressing inverter 812 that generates a voltage u(t) at its output and which voltage is in particular measured at the output of the indicated first output filter 814. The voltage u(t) is continually measured and fed back into the first microcontroller 816. The first microcontroller 816 to this extent evaluates the instantaneous values of this measured voltage u(t). These measured values, just like the output voltage, are also three-phase. However, an explanation of
The first microcontroller 816 is additionally provided with a voltage signal usetpoint,w specifying the voltage u(t) to be set according to magnitude, frequency and phase.
This setpoint value usetpoint,w is generated in the first inverter controller 818. It is dependent on a setpoint voltage Usetpoint,N and the measured values U, I, f, φ that are measured at the output of the first grid inductor 820.
The current-impressing unit 810, which operates in a current-impressing manner, has a current-impressing inverter 822 that operates in a similar manner to the voltage-impressing inverter 812 but controls to an output current i(t). This output current i(t) is recorded at the output of the indicated second output filter 824 and evaluated in the second microcontroller 826. The second microcontroller 826 is therefore provided with a current setpoint value isetpoint,w that presets the current i(t) to be generated according to magnitude, frequency and phase. The second microcontroller 826 correspondingly controls the switching actions in the current-impressing inverter 822, this being indicated by the reference sign S. Correspondingly, the first microcontroller 816 otherwise controls switching actions in the inverter 812.
The current setpoint value isetpoint,w is determined in the second inverter controller 828. It is dependent on the voltage U, the current I, the frequency f and the phase angle φ, and these variables are recorded at the output of the second grid inductor 830. The second inverter controller 828 also receives a setpoint voltage Usetpoint,N as an input variable too. The first and second inverter controllers 818 and 828 may also be combined in a common inverter controller.
The voltage-impressing unit 808 therefore ultimately generates a first current I1 and the current-impressing unit 810 ultimately generates a second current I2. These two currents I1 and I2 are added to produce the joint current IG. This flows for illustrative purposes into the symbolized farm grid 800. This is intended to be understood as an illustration because the voltage-impressing unit 808 and the current-impressing unit 810 are also part of the farm grid 800. To this extent, the joint current IG flows into the remaining part of the farm grid.
During operation, if for example a sudden change in reactive power or a phase jump occurs in the farm grid 800, this becomes noticeable in the joint current IG. Because the output current I2 of the current-impressing unit 810 is controlled by the latter, a change in the joint current IG therefore initially only leads to a change in the first current I1 of the voltage-impressing unit 808.
The change in the joint current IG has therefore initially led to a change in the first current I1, and this has been recorded by the first inverter controller 818. The first inverter controller 818 records from it, in dependence on a reactive power droop or real power droop, a new value for a voltage amplitude and/or a frequency. Correspondingly, the voltage setpoint signal usetpoint,w is adapted and transferred to the first microcontroller 816. The latter then correspondingly activates the voltage-impressing inverter 812. This correspondingly leads to a change in the voltage amplitude and/or in the frequency of the voltage, and this is measured by the current-impressing unit 810 by the measurement at the output of the second grid inductor 830 and is evaluated in the second inverter controller 828. Dependent on this, a new reactive power value and/or a new real power value is then calculated, to be specific dependent on the underlying reactive power droop or real power droop. Correspondingly, a setpoint current signal isetpoint,w is preset and transferred to the second microcontroller 826. The latter then activates the current-impressing inverter 822 correspondingly. The result is that the second current I2 now changes, as a result of which the first current I1 also changes and this in turn leads to a new adaptation by the first inverter controller 818, to be specific again based on the corresponding droops, that is to say the reactive power droop and/or the real power droop.
Ultimately, the voltage-impressing unit 808 and the current-impressing unit 810 will therefore adjust themselves to one another in such a way that they feed in a corresponding real or reactive power in a way corresponding to the droop relevant to them for the same voltage deviation and the same frequency. It has consequently been recognized that system services are known from wind power installations. However, at present these have been adapted to the physical characteristics of the previously dominant feeding-in technology, which is based on large-scale power plants with synchronous machines.
It has also been recognized that, depending on the grid operating point, in the short and medium term the feeding-in technology should adopt the system responsibility that is dominant at the time, that is to say conventional or inverter-based. Simply replicating conventional, usually slowly responding, generators by inverters does not appear to be productive.
Nevertheless, it may be necessary under some circumstances also to ensure voltage impression in the future by inverter-based systems. In this case, the inverters must not operate in a current-impressing manner, but in a voltage-impressing manner.
This has the following advantages:
An initial voltage impression can be achieved during the black start and grid recovery of an energy supply grid, while it is proposed here not to restrict the use of voltage-impressing inverters to this.
Provision of a voltage impression without additional energy stores and without an adaptation of the control strategy may become possible. DC link control by inverters on the grid side may be proposed for this.
A restriction of the frequency gradients in grids with high inverter penetration and protection of the remaining directly coupled machines becomes possible.
Static load balancing may be made possible.
Improved response to faults and handling of faults can be achieved, and thereby enhanced failure characteristics.
Provision of initial fault currents for initiating grid protection is achievable.
An instantaneous generator reaction to frequency drops and voltage drops is made possible, for example in the case of a system split.
Improved frequency and voltage stability is also achievable.
Since an inverter system with voltage impression should maintain a certain power reserve, in order that voltage-impressing operation is not put at risk by the current limits of the inverters, voltage-impressing inverters usually have a lower power density and are more expensive.
A desired aim is consequently also to provide the voltage-impressing characteristics at lowest possible cost, and consequently with fewest possible voltage-impressing inverter units.
At least some embodiments relate to open-loop and closed-loop control strategies for wind power installations and other power-electronically coupled feeding-in suppliers, in order to provide voltage-impressing characteristics in parallel grid operation.
It is intended in this way to achieve the effect of increasing the proportion of inverter-based generation in integrated grid systems and developing a closed-loop control concept for grids that are operated or realized at times almost completely by inverter-coupled feeding in of renewable energies, without any appreciable disadvantages in the reliability of the system.
The following problem in this respect was also recognized: Conventional power plants, to be specific those with directly coupled synchronous machines, have voltage-impressing characteristics in physical terms. Since most regenerative feeding-in suppliers do not have directly coupled synchronous machines, or it is not advisable for them to be used for other reasons, they will not be available in the future.
Solutions can be proposed for the following problems:
provision of voltage-impressing characteristics by inverter-fed feeding-in suppliers;
possible black start with regenerative power plant capacities;
speeded-up grid recovery with regenerative power plant capacities;
grid integration of an at times very high proportion of renewable energies in the grid;
more reliable grid operation, even in grids that are at times supplied almost completely on an inverter basis;
adoption of system responsibility in the electrical energy supply by wind power installations or wind farms;
avoidance of a pseudo-technical cap on the development of renewable energy generators;
substitution of conventional power plants; and
increasing the acceptance of regenerative feeding-in suppliers among grid operators.
According to the disclosure, at least according to one or more embodiments, the following has consequently been recognized.
Since power-electronic feeding-in suppliers usually consist of a certain number of parallel units, that is to say particularly parallel-connected converters or inverters, which exchange power with the grid, the proposed solution is to use a small number of units for the actual voltage impression and to divide the power quickly among the individual units by way of a combined quick, decentralized controller. This succeeds in making the entire feeding-in supplier act in a voltage-impressing manner with only a few units.
In a first step, the units that exchange power with the grid are divided into voltage-impressing units and current-impressing units. A ratio of for example 1:9 appears to be advisable in this respect. Therefore, such a ratio is proposed, in particular a ratio in the range from 1:5 to 1:20, preferably 1:8 to 1:16. The ratio may however also be adapted grid-situation-dependently. It relates to the rated power of each unit. In case of units of the same size, which is preferably proposed, the ratio indicates numbers, that is to say the number of voltage-impressing units to the number of current-impressing units.
For splitting the power, the current-impressing units are in this case operated with two droops:
For the reactive power in dependence on the voltage, a so-called Q(U) droop is preset. In this case, a setpoint voltage is preset for the inverter. In dependence on the deviation from the setpoint voltage, the inverter then feeds in a reactive power.
For the power in dependence on the frequency, a so-called P(f) droop is preset. In this case, at the rated frequency there will be the setpoint power, which may be a preset setpoint power or supply-related maximum power. When there is a frequency deviation, the power is adapted in accordance with a droop.
The voltage-impressing units are likewise operated with two droops, but in dependence on the other reference variable respectively. They control the voltage in dependence on the reactive power that has just been established, or the frequency in dependence on the real power that has just been established.
In this case, the two systems differ in three respects:
1. At the rated voltage and rated frequency, the operating points of the voltage-impressing inverters are in each case zero, or close to zero.
2. The droops of the current-impressing units are much steeper, to be specific approximately or at least about five times steeper, than those of the voltage-impressing units. A system deviation in the static case is therefore dealt with almost completely by the current-impressing units, because there are more current-impressing units that have a steeper droop.
3. The control of the setpoint values of the voltage-impressing units as a dependent variable is much quicker than that of the current-impressing units.
If, for example, an upward frequency drift occurs, which indicates surplus power in the grid, the voltage-impressing inverters initially take over completely the additional currents resulting from the frequency and the phase shift. This means that there is a real-power take-up, and the frequency is in the first moment “fixed”.
However, the closed-loop control quickly controls the setpoint frequency as a dependent variable to the value resulting from the droop, to be specific in accordance with the resultant power.
The current-impressing inverters then see the frequency deviation and increase the real power.
The same applies analogously to the voltage and the reactive power.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2017 112 491.8 | Jun 2017 | DE | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2018/065031 | 6/7/2018 | WO | 00 |
