Reactive-power control method for decentral voltage maintenance in a distribution network, and inverter and operating method

Abstract

The invention relates to a reactive-power control method for decentral voltage maintenance in a distribution network (2), in particular in a low-voltage network, comprising at least one decentral power generator (PV1-PV4), in which method a local mains voltage (U) of the distribution network (2) is measured at the feed point (4) of the power generator (PV1-PV4), an active power (P) is fed into the distribution network (2) and a reactive power (Q) is fed in on the basis of the local mains voltage (U), wherein the local mains voltage (U) is controlled as a controlled variable to a voltage limit value (Umax) of the local mains voltage (U), and the reactive power (Q) is used as a manipulated variable.

Description

[0001] Reactive power control methods for decentralized voltage control of a distribution network as well as inverters and operating procedures

[0002] The invention relates to a reactive power control method for decentralized voltage control of a distribution network, in particular a low-voltage network, with at least one decentralized power generator. A local grid voltage of the distribution network is detected at the feed-in point of the power generator, active power is fed into the distribution network, and reactive power is fed in depending on the local grid voltage. Further subjects of the invention are a method for operating an inverter, in particular a grid-following inverter, a decentralized power generator, in particular a photovoltaic system, a computer program, an inverter, in particular a grid-following inverter, for a decentralized power generator, in particular a photovoltaic system, and a distribution network, in particular a low-voltage network, with multiple decentralized power generators.

[0003] The power grid, which transmits and distributes electrical energy between power suppliers and consumers, is divided into different voltage levels. For example, the power grid in Germany is divided into four voltage levels with different characteristics. While the meshed power grid at the extra-high and high-voltage levels handles long-distance transmission, there is no comparable redundancy in the supply at the medium and low voltage levels. At the low-voltage level, the grid is supplied with a nominal voltage of 0.4 kV from a central transformer at the transition to the higher-level medium-voltage grid, although several parallel conductor lines can also be used. This structure enables a clear and cost-effective grid design.

[0004] Electricity grid operators must adhere to various quality criteria when connecting household customers to ensure that devices function properly and do not suffer damage. The DIN EN 50160 standard defines the requirements for voltage quality in distribution networks. These include specifications for static voltage quality, which must be met to ensure a reliable supply.

[0005] One such power quality requirement is static voltage stability. Within a one-week period, 95% of the 10-minute average values ​​must be within a voltage tolerance of 10% of the nominal voltage, for example, 230 V. However, short-term undervoltages of up to 85% of the nominal voltage are permissible.

[0006] Compliance with this voltage range, which is specified by the voltage tolerance, requires measures from the electricity supplier, especially when decentralized power generators, such as photovoltaic systems, with a large total output are connected in a grid area. The decentralized power generators cannot coordinate their respective power feed-in to the distribution grid, as there is usually no communication, especially no data exchange, between individual decentralized power generators in the distribution grid. Planning, especially centralized planning, of reactive power feed-in for all decentralized power generators while adhering to the voltage range is therefore not possible.

[0007] Compliance with the voltage range can be achieved, for example, by the decentralized power generator feeding in reactive power. Methods used for this purpose, such as the Q(U) method defined in the VDE AR-4105 standard, measure the local grid voltage of the distribution grid at the power generator's feed-in point, where it feeds in power into the distribution grid. This local grid voltage is influenced by the power feed-ins from other decentralized power generators also located in the distribution grid, which are located along the line leading from the central transformer. This local grid voltage corresponds to the voltage of the distribution grid at the feed-in point relative to the reference potential. In addition to the active power, the decentralized power generator also feeds reactive power into the distribution grid to compensate for any voltage increase caused by the active power feed-in.The strength of this reactive power fed in depends on a predefined characteristic curve, which specifies the reactive power to be fed in depending on the local grid voltage.

[0008] To feed in active and reactive power, the decentralized power generator uses its inverter, which is typically a grid-following inverter. This inverter is operated to maintain voltage in such a way that the reactive power feed-in, i.e., the reactive power feed-in into the distribution grid, occurs according to the reactive power control method. The inverter and / or the decentralized power generator either have control and / or regulation devices specifically designed for voltage control or a variably deployable computer or microcontroller loaded with a computer program designed for voltage control.

[0009] A disadvantage of these known methods, however, is that they begin feeding reactive power into the distribution grid at local grid voltages that are still far from violating the voltage range. This early reactive power feed-in results in power losses in the system and the grid, which are also associated with corresponding financial losses. These losses will increase with the increased expansion of decentralized power generators. Nevertheless, this early reactive power feed-in is necessary with the known methods, as decentralized power generators located at the beginning of the line would otherwise not participate in voltage control.It is necessary that decentralised power generators located at the beginning of the string also participate in voltage control, even though they are not yet able to detect local voltage control problems, because otherwise the decentralised power generators at the end of the string would not be able to maintain the local grid voltage at their feed-in points, which is influenced by all the decentralised power generators.

[0010] If, starting from an operating point of these processes, there is a further increase in the active power to be fed in, for example in the case of stronger solar radiation which contributes to the increase in the performance of the photovoltaic system, the local grid voltage also increases with this known process.

[0011] The uniform parameterization of a Q(U) control system for decentralized power generators, especially their inverters, which is necessary for practical reasons, leads to an insoluble dilemma between the efficiency and effectiveness of the Q(U) method, since only the voltage is taken into account. The object of the present invention is therefore to enable voltage control at the feed-in point by means of reactive power feed-in from decentralized power generators, to reduce the system and grid losses caused by reactive power, and thus to increase efficiency.

[0012] This task is solved in a reactive power control method of the type mentioned above by regulating the local grid voltage as a controlled variable to a voltage limit value of the local grid voltage and using the reactive power as a manipulated variable.

[0013] By using the voltage limit as a true reference variable of a control loop, the reactive power fed into the distribution grid is not adjusted based on a static voltage-dependent characteristic curve, but rather the reactive power feed-in is used as a reference variable to regulate the local grid voltage to the voltage limit. This prevents the specified voltage limit from being exceeded as active power feed-in increases. This reactive power control method can reduce the system and grid losses caused by reactive power by 70-90% compared to the conventional method, without reducing the effectiveness with regard to the integration potential, i.e., efficiency is increased.The integration potential is a measure of the maximum output of all decentralized power generators that can be connected to the distribution grid without violating the voltage tolerance or load capacity of transformers and lines. By reducing system and grid losses, significant economic savings can also be achieved for the grid operators and the power generator. The local grid voltage is to be controlled to the voltage limit as a reference variable for the reactive power control process.

[0014] With the reactive power control method, the voltage of the distribution grid is maintained decentrally. A central component responsible for voltage control of the entire distribution grid is dispensed with. Individual decentralized power generators in the distribution grid can independently contribute to voltage control. External instructions, control and / or regulation commands from a central component of the distribution grid directed to at least one decentralized power generator can be avoided. The regulation of the local grid voltage to the voltage limit can be performed decentrally, in particular by the power generator at whose feed-in point the local grid voltage is measured.

[0015] Preferably, the method is implemented serverlessly, in particular in the inverter of the decentralized power generator. A central server that serves for data exchange between components of the distribution grid, acts as a central control device that directly or indirectly controls individual or multiple power generators, acts as a central control device that directly or indirectly controls individual or multiple power generators, and / or on which the reactive power control process is implemented, can be dispensed with.

[0016] Advantageously, particularly with regard to the necessary infrastructure and the susceptibility to errors in voltage regulation, no data on grid parameters, in particular no exchange of data on measured values, impedance, or local impedance of the distribution grid, takes place between components of the distribution grid, especially between power generators. Voltage control with this reactive power control method can be achieved, in particular, without communication. The means of the decentralized power generator used to implement this reactive power control method cannot be involved in any data exchange with corresponding means of other decentralized power generators in the distribution grid and / or a central control or regulation device of the distribution grid.

[0017] Preferably, the local grid voltage is recorded locally, especially exclusively, at the feed-in point. Recording by a central component of the distribution grid and, in particular, error-prone back-calculation to the local grid voltage from non-local grid parameters of the distribution grid can be avoided. The local grid voltage can, in particular, be measured locally. The local recording of the local grid voltage can be carried out by the decentralized power generator at whose feed-in point the local grid voltage is recorded.

[0018] Advantageously, the method uses only locally recorded values ​​for grid parameters such as the local grid voltage, reactive power, and / or active power. The use of non-locally recorded values, particularly those determined by back-calculating local values ​​from these values, which may deviate from the actual local values ​​at the feed-in point, can be avoided.

[0019] Preferably, a control deviation is determined from the local grid voltage and the voltage limit. This control deviation can be used as an input variable for a control system. Based on the control deviation, the control system can determine the reactive power that the decentralized power generator should feed into the distribution grid. The reactive power can thus serve as a control variable for the control loop.

[0020] It is also advantageous if the reactive power is only fed in when the voltage limit of the local grid voltage is reached. In this case, it is possible for no reactive power to be fed in if the detected local grid voltage is lower than the voltage limit. In this way, the local grid voltage can be maintained at the voltage limit when the voltage limit is reached. The participation of decentralized power generators located at the beginning of the string in the voltage control of decentralized power generators downstream of the string, without there being a local voltage band violation at their feed-in points, can be limited to times with a generally high power feed-in to the distribution grid. The voltage limit can form an upper limit of a voltage dead band, within which unnecessary reactive power feed-in can be prevented.The voltage deadband can reduce losses in both the inverter and the grid operator's lines.

[0021] In a further embodiment of the invention, the voltage limit is a variable limit dependent on the active power feed-in. The voltage limit, as a reference variable, can assume a value that changes with the active power fed in. The variable voltage limit can be determined at the feed-in point as a function of the active power feed-in, i.e., the active power fed in. In this way, the size of the voltage deadband can be changed depending on the power. By designing the voltage limit as an active-power-dependent variable limit, the range of the local grid voltage within which reactive power is fed into the distribution grid can be specified for different active powers to be fed in.

[0022] In this context, it is particularly advantageous if the variable voltage limit is reduced depending on the active power. At higher active power levels, the feed-in of reactive power can occur at lower local grid voltages than would be the case with lower active power levels. In this way, the reactive power feed-in in the range of lower active power levels can only occur at significantly higher local grid voltages. The voltage deadband can be reduced as the active power feed-in increases by reducing the variable voltage limit depending on the active power. The reactive power feed-in and the associated losses can be avoided over a wider range of the local grid voltage in the range of lower active power levels than with higher active power levels.

[0023] It is further advantageous if the variable voltage limit is determined using a voltage limit / active power feed-in characteristic curve, in particular a continuous and / or monotonically decreasing one, depending on the active power feed-in. With a voltage limit / active power feed-in characteristic curve, the values ​​of the variable voltage limit can be easily specified for different active power feed-ins. With a continuous voltage limit / active power feed-in characteristic curve, sudden changes in the control process can be avoided. With a monotonically decreasing voltage limit / active power feed-in characteristic curve, an increase in the variable voltage limit with increasing active power feed-in can be prevented, since the voltage limit decreases with increasing active power feed-in or remains at most constant over an active power range.

[0024] It is further advantageous if the voltage limit-active power feed-in characteristic curve has a power threshold, above which the relationship between the active power feed-in and the voltage limit changes. By using a power threshold, the voltage limit-active power feed-in characteristic curve can be divided into at least two different regimes. In these regimes, the active power-dependent change behavior of the voltage limit can differ from one another. The power threshold can be in the range of 20% to 80%, in particular in the range of 50% to 70%, preferably in the range of 55% to 65%, particularly preferably at 60%, of the nominal power of the power generator.

[0025] In this context, it is particularly advantageous if the variable voltage limit assumes a constant value for active power feed-ins below the power threshold. In this way, a voltage deadband of constant magnitude can be specified for active power feed-ins below the power threshold. The constant value can be in the range of 106% to 110%, in particular in the range of 107% to 109%, preferably in the range of 107.5% to 108.5%, particularly preferably at 108%, of the nominal voltage of the distribution grid.

[0026] In a further embodiment of the invention, the variable voltage limit decreases, particularly linearly, with increasing active power feed-in for active power feed-in above the power threshold. As the active power feed-in increases above the power threshold, the voltage deadband can thus shrink, so that the reactive power feed-in occurs at increasingly lower local grid voltages.

[0027] The variable voltage limit preferably assumes a minimum value when the active power feed-in corresponds to the rated power of the power generator. The rated power of the power generator can be the nominal active power, the maximum that the power generator can achieve. When the rated power is reached, the maximum possible active power feed-in can occur by the power generator. The minimum value assumed by the variable voltage limit at the rated power can be a predeterminable voltage offset below the constant value assumed by the variable voltage limit at low active power feed-ins below the power threshold. The minimum value can be in the range from 103% to 107%, in particular in the range from 104% to 106%, preferably in the range from 104.5% to 105.5%, particularly preferably at 105%, of the rated voltage of the distribution grid.

[0028] According to a further embodiment of the invention, it is proposed that a maximum reactive power is specified up to which the reactive power to be fed in can be increased. Exceeding the variable voltage limit can be prevented up to this maximum reactive power. The fed-in reactive power can reach saturation above the maximum reactive power. The maximum reactive power can, in particular, correspond to a maximum displacement factor of 0.90. After the maximum reactive power is reached, the control can remain active but initially no longer influence the manipulated variable and thus the local grid voltage. Only after a drop in the local grid voltage or a reduction in the active power feed-in, in particular due to influences outside the control system, can the control system influence the local grid voltage again.

[0029] In this context, it has proven advantageous if, once the maximum reactive power is reached, it is fed into the grid as reactive power if the voltage limit is exceeded. Once the maximum reactive power is reached, the reactive power control method cannot cause a further increase in reactive power feed-in, especially if the local grid voltage and / or active power feed-in increase. Once the maximum reactive power is reached, the decentralized power generator can continuously feed the maximum reactive power into the distribution grid, regardless of a further increase in the local grid voltage.

[0030] A further embodiment of the invention provides for only inductive reactive power to be fed in. This eliminates the need for capacitive reactive power. This prevents unwanted voltage increases caused by capacitive reactive power.

[0031] In an advantageous embodiment of the invention, a control loop without a steady-state setpoint deviation is used. A control loop without a steady-state setpoint deviation can exhibit no control deviation or control difference between the controlled variable and the specified reference variable in the steady-state state. This can be achieved, in particular, with a control loop exhibiting integrating control behavior. Such a control loop can contain at least one pure integrator. In particular, the control loop can be configured as an I, PI, or PID controller.

[0032] According to one embodiment of the invention, the control is carried out in discrete time steps. However, the control can also be carried out continuously.

[0033] The reactive power feed-in can be calculated in particular using the equation

[0034] Qt+1 = Qt + K ( Umess, t - Umax). Here, Q t+i is the reactive power to be fed in at time t+1, which is determined from the reactive power Qt fed in at time t, a weighting factor K specified for the transient response over time and the difference between the local grid voltage Umess, t measured at time t and the voltage limit value Umax.

[0035] When using a variable voltage limit Umax(P) dependent on the active power feed-in, the formula

[0036] Qt+1 = Qt + K (Urness,t - Umax(P) ) can be used.

[0037] If only inductive reactive power is to be fed in, the formula

[0038] Qt+1 = max(Qt + K ( Umess,t - Umax(P)), 0) can be used, in which only an inductive reactive power, ie when using the consumer metering arrow system a reactive power with a positive value, is fed in.

[0039] In a method for operating an inverter of the type mentioned at the outset, it is proposed to solve the above-mentioned problem that a reactive power feed into a distribution network for decentralized voltage control is controlled in the manner described above, thereby resulting in the advantages described in connection with the reactive power control method.

[0040] In the case of the computer program of the type mentioned at the outset, in order to achieve the above-mentioned object, it is proposed that the program comprises instructions which, when the program is executed by a computer, cause the computer to carry out the reactive power control method described above, thereby resulting in the advantages described in connection with the reactive power control method.

[0041] The computer program can be executed by a computer controlling the decentralized power generator, in particular an inverter of the decentralized power generator. The decentralized power generator, in particular its inverter, can have detection means for detecting the local grid voltage. The computer program controlling the decentralized power generator can read the detected local grid voltage and / or cause the detection means to detect the local grid voltage.

[0042] To achieve the above-mentioned object, a computer-readable medium is also proposed on which the above-described computer program is stored, thereby resulting in the advantages described in connection with the computer program. In an inverter of the type mentioned above, to achieve the above-mentioned object, it is proposed that the inverter comprise means, in particular a control device, for decentralized voltage control in the manner described above, in particular for executing the above-described computer program.

[0043] The features described in connection with the method for operating an inverter according to the invention can also be applied to the inverter individually or in combination. This results in the same advantages as those already described.

[0044] A control device can define a control loop in a fixed circuit, in particular, comprising a hardware implementation of the control loop, such as a controller. The control device can have a memory for storing a voltage limit or the voltage limit-active power feed-in characteristic curve. The control device can have a readout device for reading the active power-dependent value of the variable voltage limit from the voltage limit-active power feed-in characteristic curve.

[0045] Alternatively or in addition to a control device, the means for maintaining the voltage may also be a computer or a programmable computer chip installed in the inverter, on which the previously described computer program is stored and can be executed for voltage maintenance.

[0046] The inverter may have a detection device for detecting the local grid voltage.

[0047] In a distribution network of the type mentioned at the outset, it is proposed to achieve the above-mentioned object that said network has means, in particular control devices, for decentralized voltage control in the manner described above, in particular inverters of the type described above.

[0048] The features described in connection with the inventive reactive power control method for decentralized voltage control and / or the inventive inverter can also be applied individually or in combination to the distribution grid. This results in the same advantages as those already described.

[0049] Each of the decentralized power generators can be assigned a means for decentralized voltage control. Using this means, each of the decentralized power generators can contribute to voltage control locally and decentrally. The means, and in particular the decentralized power generators, cannot exchange data with each other. In particular, the means for decentralized voltage control can be part of the decentralized power generator.

[0050] A control device can define a control loop in a fixed circuit, in particular, comprising a hardware implementation of the control loop, such as a controller. The control device can have a memory for storing a voltage limit or the voltage limit-active power feed-in characteristic curve. The control device can have a readout device for reading the active power-dependent value of the variable voltage limit from the voltage limit-active power feed-in characteristic curve.

[0051] The respective voltage control means may, alternatively or in addition to a control device, also be a computer or a programmable computer chip installed in the respective decentralized power generator, on which the previously described computer program is stored and executable for voltage control. Further details and advantages of a reactive power control method according to the invention, a method according to the invention for operating an inverter, a computer program according to the invention, an inverter according to the invention, and a distribution network according to the invention will be explained below using the exemplary embodiments of the invention schematically illustrated in the figures. Therein:

[0052] Fig. 1 : the strand of a distribution network with several decentralized

[0053] power generators,

[0054] Fig. 2: the control loop of the reactive power control method according to the invention,

[0055] Fig. 3: a voltage limit active power feed-in characteristic curve and

[0056] Fig. 4: the control behavior of the reactive power control with increasing active power feed-in.

[0057] Fig. 1 shows a distribution network 2 in the form of a low-voltage network, which is connected to the higher-level medium-voltage network 200 via a central transformer 100. Starting from the central transformer 100, the distribution network 2 extends as a line 3.

[0058] Along this string 3, the distribution grid 2 has several decentralized power generators PV1 to PV4, which are arranged between the string start 3.1 and the string end 3.2. These decentralized power generators PV1 to PV4 are primarily photovoltaic systems. These decentralized power generators PV1 to PV4 generate electricity at various points in the distribution grid 2 and, together with this electricity, feed an active power P into string 3 of the distribution grid 2 via their respective feed-in points 4. This additional and generally temporally non-constant power feed into the distribution grid 2 would, without correction, lead to significant location-dependent fluctuations in the voltage of the distribution grid 2.The voltage quality required to ensure a stable power supply to the end consumers also located in distribution network 2 could not be maintained in this way, since the local voltage of distribution network 2 could not be maintained within a voltage band defined by predetermined voltage tolerances, typically + / - 10%, around a nominal voltage.

[0059] In order to keep the local voltages within the limits of the voltage range, the decentralized power generators PV1 to PV4 measure the local grid voltage U of the distribution grid 2 at their respective feed-in points 4 in order to achieve decentralized voltage control on this basis. The individual decentralized power generators do not exchange data with each other or with a central control device of the distribution grid 2. To achieve this decentralized voltage control, the decentralized power generators PV1 to PV4 feed not only an active power P but also a reactive power Q, which depends on the local grid voltage U, into the distribution grid 2 at their feed-in points 4.

[0060] Fig. 2 shows a control loop 1 with which the reactive power control method according to the invention for decentralized voltage control of the distribution network 2 is carried out. In the following, the control loop 1 is described as the control loop of the decentralized power generator PV4, although the other power generators PV1 to PV3 of the distribution network 2 can also have control loops 1 designed in the same way. The local grid voltage U is used as the controlled variable, which is recorded as described above and controlled by the control loop 1 for voltage control. In addition to the local grid voltage U as the controlled variable, the control loop 1 uses a voltage limit value Umax as the reference variable to which the local grid voltage U is to be regulated.

[0061] In the next step, a control deviation E is determined from the local grid voltage U and the voltage limit value Umax and used as an input variable for control device 6. Based on the control deviation E, control device 6 determines the reactive power Q, which is fed into the distribution grid 2 by an inverter of the decentralized power generator PV4 and thus serves as a control variable for control loop 1.

[0062] Based on the control deviation E, which corresponds to the difference between the local grid voltage U and the voltage limit Umax, the control device 6 can determine whether the local grid voltage U is less than the voltage limit Umax. If the local grid voltage U is below the voltage limit Umax, the control device 6 can ensure that no reactive power Q is fed into the distribution network 2. In this way, the reactive power Q is only fed in when the voltage limit Umax of the local grid voltage U is reached.

[0063] Along the control system 7 of the control loop 1, the fed-in reactive power Q causes a change in the local grid voltage U through the control behavior 8. In addition to the influence of the reactive power Q fed in by the decentralized power generator PV4, the active power feed-ins and reactive power feed-ins of the other decentralized power generators PV1 to PV3 as well as other influences also influence the local grid voltage U at the feed-in point 4 of the decentralized power generator PV4. These influences on the local grid voltage U, which are not directly caused by the decentralized power generator PV4, are summarized as the disturbance behavior 9 of the control system 7, which affects the local grid voltage U as the disturbance variable Z.

[0064] Both the disturbance variable Z and the fed-in reactive power Q combine to cause a change in the local grid voltage U at feed-in point 4 of the decentralized power generator PV4. This changed local grid voltage U is then recorded and compared again with the voltage limit Umax during the next pass through control loop 1. This reactive power control can be implemented either in discrete time steps or as a continuous-time control.

[0065] To ensure that after control loop 1 has settled over time, there is no longer any steady-state setpoint deviation, i.e., that the local grid voltage U in the controlled state assumes the voltage limit Umax serving as the reference variable without deviation, control device 6 can exhibit integrating control behavior. For this purpose, control device 6, and thus also control loop 1, contains a pure integrator. It can be configured as an I, PI, or PID controller.

[0066] Although the voltage limit Umax can be specified as a power-independent limit, it has proven particularly advantageous if the voltage limit Umax is a variable limit dependent on the fed-in active power P, i.e. the active power feed-in P, and in particular is reduced depending on the active power. To determine the size of this variable voltage limit Umax, a voltage limit-active power feed-in characteristic curve 5 is used, as shown, for example, in Fig. 3. This continuously and monotonically decreasing voltage limit-active power feed-in characteristic curve 5 indicates the value of the voltage limit Umax as a function of the fed-in active power P of the decentralized power generator PV4. As already shown in Fig.In Fig. 2, the voltage limit active power feed-in characteristic curve 5 is described as a characteristic curve for controlling the decentralized power generator PV4, although the other power generators PV1 to PV3 of the distribution network 2 can also be controlled using the same voltage limit active power feed-in characteristic curve 5. The voltage limit Umax is specified in Fig. 3 as a multiple of the nominal voltage of the distribution network 2, which corresponds, for example, to 230 V. The fed-in active power P is specified as a multiple of the nominal power of the decentralized power generator PV4, whereby the nominal power is the maximum active power P that the decentralized power generator PV4 can feed into the distribution network 2 under ideal conditions.

[0067] The voltage limit active power feed-in characteristic curve 5 has a power threshold P reiwhich divides the voltage limit active power feed-in characteristic curve 5 into two areas with different dependencies of the voltage limit Umax on the active power feed-in P. In the example shown in Fig. 3, this power threshold Prei is 60% of the nominal power of the decentralized power generator PV4.

[0068] If the fed-in active power P is below the power threshold Prei, the variable voltage limit Umax has a constant value Umax,i. In the example in Fig. 3, this constant value Umax,i is 108% of the nominal voltage of distribution grid 2. As long as the decentralized power generator PV4 feeds active power P below the power threshold Prei into distribution grid 2, no additional reactive power Q is fed into distribution grid 2, provided the local grid voltage U is below this constant value Umax,i, which is relatively high compared to the known methods. Losses associated with the reactive power feed-in Q can thus be prevented across this parameter range.

[0069] For active power feed-ins P above the power threshold P reiThe variable voltage threshold Umax decreases linearly with increasing active power feed-in P until it reaches a minimum value Umax, 2 at the active power feed-in P that corresponds to the nominal power (P=1 ) of the decentralized power generator PV4. In the active power range between the power threshold P rei and the rated power of the decentralized power generator PV4, the variable voltage limit value Umax decreases from the value Umax,i by the voltage offset Uoffset to its minimum value Umax,2.

[0070] Fig. 4 shows the control behavior of the reactive power control with increasing active power feed-in P as an example for the distribution network 2 shown in Fig. 1 with four decentralized power generators PV1 to PV4, whereby each of the decentralized power generators PV1 to PV4 is controlled via its own control loop 1 according to the reactive power control method described above. The local grid voltage U is plotted at the respective feed-in points 4 of the decentralized power generators PV1 to PV4. The local grid voltage U is given as a multiple of the nominal voltage of the distribution network 2. The fed-in active power P of the individual decentralized power generators PV1 to PV4 is shown, as already in Fig. 3, as a factor of the nominal power of the respective decentralized power generators PV1 to PV4.

[0071] For a better understanding of the control behavior, the voltage limit active power feed-in characteristic curve 5 indicating the active power-dependent variable voltage limit Umax according to Fig. 3 is shown as a dashed line in Fig. 4.

[0072] Initially, the local grid voltages U at the feed-in points 4 of the decentralized power generators PV1 to PV4 increase linearly with increasing active power feed-in P. As long as the active power feed-in P has not yet reached the power threshold value Prei and the local grid voltages U have not yet reached the voltage limit active power feed-in characteristic curve 5, the curves are located in a comparatively wide voltage dead band B, in which no reactive power Q is fed into the distribution grid 2.

[0073] From the power threshold P reithe width of this voltage dead band B decreases with increasing active power P until the voltage dead band B reaches its narrowest width at the rated power of the decentralized power generators PV1 to PV4.

[0074] Even after exceeding the power threshold P reithe curves of the four power generators PV1 to PV4 are initially still below the voltage limit active power feed-in characteristic curve 5, so that initially no reactive power is fed in. At point S1, the local grid voltage U at the feed-in point 4 of the decentralized power generator PV4 reaches the voltage limit Umax applicable for this active power P, so that the decentralized power generator PV4 begins to feed in reactive power Q to maintain the voltage, i.e. to maintain the voltage band of the distribution grid 2. Since the local grid voltages U of the other decentralized power generators PV1 to PV3 have not yet reached the voltage limit Umax with this power feed-in P, the decentralized power generators PV1 to PV3 are not yet feeding any reactive power Q into the distribution grid 2.

[0075] As the active power feed-in P increases, the controlled decentralized power generator PV4 feeds increasingly more reactive power Q into the distribution network 2, which leads to a reduction in the local grid voltages U of all decentralized power generators PV1 to PV4 without violating the specified voltage range of the distribution network 2.

[0076] The increase in reactive power feed-in Q of the decentralized power generator PV4 continues with increasing active power feed-in P, until at point S2 a maximum reactive power specified by the inverter manufacturer is fed into distribution grid 2 by the decentralized power generator PV4. Reactive power exceeding this maximum reactive power cannot be fed into distribution grid 2 by the decentralized power generator PV4. As the active power feed-in P continues to increase, the decentralized power generator PV4 continues to feed only the maximum reactive power into distribution grid 2, even if the voltage limit Umax is exceeded.

[0077] Since the reactive power Q fed in by the decentralized power generator PV4 does not increase any further from point S2, the local grid voltages U at the feed-in points 4 of the decentralized power generators PV1 to PV4 increase again with increasing active power feed-in P.

[0078] At point S3, the local grid voltage U at feed-in point 4 of the decentralized power generator PV3 finally reaches the voltage limit Umax applicable for this active power P, so that the decentralized power generator PV3 begins to feed in reactive power Q to maintain the voltage. As the power feed-in P increases, this voltage maintenance by the decentralized power generator PV3 also causes the local grid voltages U to decrease until the decentralized power generator PV3 also feeds its maximum reactive power into the distribution grid 2 at point S4. As the active power feed-in P continues to increase, the decentralized power generator PV3 also only feeds its maximum reactive power into the distribution grid 2, so that the local grid voltages U increase again as the active power feed-in P increases.

[0079] The increase in the local grid voltage U with increasing active power feed-in P continues until the local grid voltage U at the feed-in point 4 of the decentralized power generator PV2 reaches the active power-dependent voltage limit Umax at point S5. From point S5, the decentralized power generator PV2 begins to feed reactive power Q into the distribution grid 2 until, with further increasing active power feed-in P, this reactive power at point S6 assumes the value of the maximum reactive power that the decentralized power generator PV2 can feed into the distribution grid 2. As the active power feed-in P continues to increase, the decentralized power generator PV2 continues to feed this maximum reactive power into the distribution grid 2 as reactive power Q.Since the power generators PV1 to PV4 each feed only inductive reactive power Q into the distribution network 2, the local grid voltage U decreases during the phases in which the reactive power feed-in Q increases and does not increase, as would be the case if capacitive reactive power Q were fed in.

[0080] As can be seen in Fig. 4, the described reactive power control method can prevent the local grid voltage U at one of the feed-in points 4 of the decentralized power generators PV1 to PV4 from violating the specified voltage band of the distribution grid 2, even if the power generator PV1 to PV4 can no longer increase the reactive power Q it feeds into the distribution grid 2. Even the local grid voltage U of the downstream decentralized power generator PV4 does not exceed the limit of 108% of the nominal voltage of the distribution grid 2 for any active power P fed in and is permanently within a 10% voltage band around the nominal voltage of the distribution grid 2.

[0081] Using the reactive power control method described above in the method for the corresponding operation of an inverter, as well as such an inverter, it is possible to enable voltage control at a feed-in point 4 of a decentralized power generator PV1 to PV4 and to reduce the system and grid losses caused by reactive power Q. Reference symbols:

[0082] 1 control loop

[0083] 2 Distribution network

[0084] 3 strand

[0085] 3.1 Strand beginning

[0086] 3.2 Strand end

[0087] 4 Feed-in point

[0088] 5 Voltage limit active power feed-in characteristic curve

[0089] 6 Control device

[0090] 7 Control system

[0091] 8 Positioning behavior

[0092] 9 Disruptive behavior

[0093] B Voltage deadband

[0094] E Control deviation

[0095] P Active power

[0096] Price-performance threshold

[0097] PV1 -PV4 power generators

[0098] Q. Reactive power

[0099] S1 -S6 point

[0100] U local mains voltage

[0101] Umax voltage limit

[0102] Umax,1 constant value

[0103] Umax, 2 minimum value

[0104] Uoffset voltage offset

[0105] Z disturbance variable

[0106] 100 central transformer

[0107] 200 medium-voltage network

Claims

Patent claims:

1. Reactive power control method for decentralized voltage control of a distribution network (2), in particular a low-voltage network, with at least one decentralized power generator (PV1 -PV4), wherein a local network voltage (U) of the distribution network (2) is detected at the feed-in point (4) of the power generator (PV1 -PV4), an active power (P) is fed into the distribution network (2) and a reactive power (Q) is fed in depending on the local network voltage (U), characterized in that the local network voltage (U) is controlled as a control variable to a voltage limit value (Umax) of the local network voltage (U) and the reactive power (Q) is used as a manipulated variable.

2. Reactive power control method according to claim 1, characterized in that the reactive power (Q) is only fed in when the voltage limit (Umax) of the local grid voltage (U) is reached.

3. Reactive power control method according to one of claims 1 or 2, characterized in that the voltage limit (Umax) is a variable limit dependent on the active power feed-in (P).

4. Reactive power control method according to claim 3, characterized in that the variable voltage limit (Umax) is reduced depending on the active power.

5. Reactive power control method according to one of claims 3 to 4, characterized in that the variable voltage limit (Umax) is determined by means of a voltage limit-active power feed-in characteristic curve (5), in particular a continuous and / or monotonically decreasing one. is determined by the active power feed-in (P).

6. Reactive power control method according to claim 5, characterized in that the voltage limit-active power feed-in characteristic curve (5) includes a power threshold value (P rei) exhibits, above which the dependence between the active power input (P) and the voltage limit (Umax) changes.

7. Reactive power control method according to claim 6, characterized in that, in the case of active power feed-ins (P) below the power threshold (Prei), the variable voltage limit (Umax) assumes a constant value (U max,1 ).

8. Reactive power control method according to one of claims 6 or 7, characterized in that in the case of active power injections (P) above the power threshold (P) re i) the variable voltage limit (Umax) decreases with increasing active power input (P), in particular linearly.

9. Reactive power control method according to one of claims 3 to 8, characterized in that the variable voltage limit (Umax) assumes a minimum value (Umax, 2) at an active power feed-in (P) corresponding to the rated power of the generator (PV1 -PV4).

10. Reactive power control method according to one of the preceding claims, characterized in that a maximum reactive power is specified up to which the reactive power (Q) to be fed in can be increased.

11. Reactive power control method according to claim 10, characterized in that, from the moment the maximum reactive power is reached, this is reduced at A reactive power control method according to one of the preceding claims, characterized in that only inductive reactive power (Q) is fed in. A reactive power control method according to one of the preceding claims, characterized in that a control loop (1) without steady-state setpoint deviation is used. A reactive power control method according to one of the preceding claims, characterized in that the control is carried out in discrete time steps. A method for operating an inverter, in particular a grid-following inverter, of a distributed power generator (PV1-PV4), in particular a photovoltaic system, in which a reactive power feed-in (Q) into a distribution network (2) is controlled for distributed voltage control according to claim 1.A computer program comprising commands which, when executed by a computer, cause the computer to execute the method according to claim 1. A computer-readable medium on which the computer program according to claim 16 is stored. An inverter, in particular a grid-following inverter, for a distributed power generator (PV1-PV4), in particular a photovoltaic system, characterized by… that this means, in particular a control device, for decentralized voltage control according to claim 15, in particular for executing the computer program according to claim 16.

19. Distribution network (2), in particular low-voltage network, with several decentralized power generators (PV1 - PV4), which have means, in particular control devices, for decentralized voltage control according to claim 1, in particular inverters according to claim 18.

Images