METHOD FOR OPERATING A POWER SUPPLY PLANT, PLANT CONTROLLER FOR A POWER SUPPLY PLANT AND POWER SUPPLY PLANT

Abstract

Method for operating a power supply system (10) which is connected to an AC supply network (12) via a transformer (14) and exchanges electrical power with the AC supply network (12) via the transformer (14), wherein the transformer (14) is connected on a first side to the AC supply network (12) and on a second side to an AC system network (18) of the power supply system (10), wherein the power supply system (10) has at least one inverter (22, 24) which exchanges electrical power between a DC unit (26, 28) on the DC side of the inverter (22, 24) and the AC system network (18) on the AC side of the inverter (22, 24), comprising the steps: • Receiving a vector length (V) of the at least one inverter (22, 24) by a system controller (20), wherein the vector length (V) is determined in the at least one inverter (22, 24) and includes the ratio of the amplitudes of the voltages on the AC side and on the DC side of the at least one inverter (22, 24), • Setting an AC setpoint voltage for the AC system network (18) by the system controller (20) as a function of the vector length (V), wherein the AC setpoint voltage is reduced if the vector length (V) exceeds a first predefinable limit, and wherein the AC setpoint voltage is increased if the vector length (V) falls below a second predefinable limit, the second limit being below the first limit, • Transmitting the AC setpoint voltage to the transformer (14), wherein a tap changer (16) of the transformer (14) is arranged to set a turns ratio (T) such that the product of the voltage in the AC supply network (12) and the turns ratio approximates the AC setpoint voltage, • wherein the distance between the first and second limit depends on the step width of the tap changer (16) and is in particular proportional to the relative voltage change of at least 3 switching steps of the tap changer.

Description

Technical field of the invention

[0001] Power supply systems exchange electrical power with a higher-level power supply network via a grid connection point. Systems with power electronic converters, such as photovoltaic (PV) systems, energy storage systems, or large consumers like electrolyzers, are typically connected to the higher-level network, e.g., an AC power supply network, via a transformer. These large systems may have a system controller that, among other things, adjusts the apparent power, consisting of active and reactive power, at the grid connection point. State of the art

[0002] EP 2 504 901 A2 describes a system controller for a PV system which determines an AC target voltage as a function of the DC voltages of all inverters of the PV system and sets a transformation ratio of the transformer at the grid connection point as a function of this.

[0003] In EP 3 429 051 A1, the efficiency of an inverter is influenced by a change in the transformation ratio of the transformer at the grid connection point.

[0004] CN 104 104 103 A and CN 103 986 181 A describe how to adjust the transformation ratio of a transformer depending on the MPP voltage (Maximum Power Point Voltage) of a PV system.

[0005] The rated power of a power converter in a power supply system, particularly an inverter, is generally limited by the maximum current of the currents converted within the inverter, especially by the maximum AC-side grid current. For a given (AC) current, the rated power of the inverter can be increased by increasing the (AC) voltage. Thus, the rated power of the inverter increases linearly with increasing AC voltage. Therefore, for technical and economic reasons, it is advantageous to be able to select the highest possible rated AC voltage.

[0006] The higher the nominal AC voltage, the higher the minimum DC voltage at the inverter's DC input must be to maintain regulated operation. In particular, for a known inverter with a bridge inverter circuit and a DC link directly connected to a DC voltage source or sink, the minimum DC voltage is the voltage required to reliably exchange an AC current with the desired apparent power with the AC supply network at a given AC voltage.

[0007] Provided that the overarching AC supply network is considered rigid and the transformation ratio of the transformer between the inverter and the AC supply network is considered fixed, the fixed point of the AC voltage in the power supply system is often the voltage at the grid connection point, i.e., the voltage at the AC supply network-side winding of the transformer.

[0008] The AC voltage on the AC side of the inverter is influenced by various factors, so that, depending on the size of the system and the corresponding number of installations between the grid connection point and the inverter, significant changes in the AC voltage can occur in the power supply system. The AC voltage can dynamically depend on the apparent power of the inverter, whereby, in particular, reactive power demanded by the AC supply network in the overexcited range can noticeably increase the AC voltage at the inverter due to the inductive properties of the installations, especially the transformer itself.

[0009] A tap changer on the transformer at the grid connection point allows the transformation ratio between the AC voltage at the transformer's power supply and the AC voltage in the AC supply network to be adjusted. Typically, such a tap changer is, for example, model [model number missing in original text]. Fig. 5492 Tap changer type UCG, operated by its own control unit in such a way as to counteract changes in the voltage of the AC supply network in order to decouple the AC voltage within the power supply system from such changes.

[0010] US 2010 / 0 277 002 A1 discloses a power conversion system comprising a power converter that converts variable DC power into AC power at a first voltage, a transformer that transforms the AC power into a second voltage, and a tap changer that sets the second voltage to a set voltage.

[0011] From DE 10 2018 203 889 A1, an energy storage system is known comprising an energy storage device, a power converter for converting the DC voltage of the energy storage device into an AC voltage, and a transformer for transforming the AC voltage of the power converter into a mains voltage, in which a control device adjusts the transformation ratio of the transformer depending on the DC voltage applied to the energy storage device.

[0012] DE 20 2015 101 806 U1 discloses a renewable energy plant with an energy generation unit and a transformer, the primary side of which is connected to the energy generation unit and the secondary side of which is connected to a power grid, wherein a plant control system can adjust the transformation ratio of the plant transformer depending on a multitude of possible parameters of the energy plant and / or the energy generation unit.

[0013] In this application, the abbreviation DC (English: direct current) stands for direct current or direct voltage, depending on the context, and the abbreviation AC (English: alternating current) accordingly stands for alternating current or alternating voltage. Object of the invention

[0014] The invention is based on the objective of providing a method for operating an energy supply system, a system controller for an energy supply system, and an energy supply system that enables more efficient and / or more flexible operation. Solution

[0015] The problem is solved by a method having the features of independent claim 1. The problem is further solved by a method having the features of independent claim 5, a system controller having the features of independent claim 12, a system controller having the features of independent claim 13, a power supply system having the features of claim 14, and a power supply system having the features of claim 15. Advantageous embodiments of the methods are claimed in the dependent claims. Description of the invention

[0016] A method according to the invention serves to operate a power supply system which is connected to an AC supply network via a transformer and exchanges electrical power with the AC supply network via the transformer. The transformer is connected on one side to the AC supply network and on the other side to an AC system network of the power supply system. The power supply system includes an inverter which exchanges electrical power between a DC unit on the DC side of the inverter and the AC system network on the AC side of the inverter. The method comprises the following steps: • Receiving a vector length from the inverter by a plant controller of the power supply system, wherein the vector length is determined in the inverter and includes the ratio of the amplitudes of the voltages on the AC side and on the DC side of the inverter, • Setting an AC setpoint voltage for the AC system network as a function of the vector length, wherein the AC setpoint voltage is reduced if the vector length exceeds a first predefinable limit, and wherein the AC setpoint voltage is increased if the vector length falls below a second predefinable limit, the second limit being below the first limit, • Transmitting the AC setpoint voltage to the transformer, wherein a tap changer of the transformer is provided to set a turns ratio such that the product of the voltage in the AC supply network and the turns ratio approximates the AC setpoint voltage, • wherein the distance between the first limit value and the second limit value depends on the step width of the tap changer and is in particular proportional to the relative voltage change of at least 3 switching steps of the tap changer.

[0017] The energy supply system can be, in particular, a photovoltaic (PV) system designed primarily for feeding electrical power into the AC power grid. Such a PV system can comprise a large number of electrical components, especially photovoltaic (PV) modules, distributed decentrally over a large area. A group of PV modules connected in a string, i.e., in series, is also called a photovoltaic string (PV string). A PV generator can have one or more PV strings connected in parallel. A DC unit can, for example, have one or more such PV strings or one or more such PV generators and be connected to the inverter. An extensive PV system with a total output in the megawatt range can have a large number of DC units and correspondingly extensive installations such as AC lines and, if necessary,have intermediate transformers.

[0018] The DC unit can also be, for example, a DC storage unit that can be charged and discharged using DC current, allowing the power supply system to draw power from or feed power into the AC grid as needed. A DC unit can also be a DC load, such as an electrolysis unit or similar device, in which case the power supply system may be primarily designed to draw power from the AC grid.

[0019] The inverter of the power supply system sets a DC voltage on its DC side, which can be aligned with the desired operating point of the DC unit, for example, the maximum power point (MPP) of a PV generator. If the PV generator is operating at a relatively high DC voltage at the MPP, it is desirable to feed the PV power into the grid at the highest possible AC voltage to keep the inverter's output currents and the currents in the system as low as possible, thus minimizing losses. However, if the MPP voltage drops, for example, due to low solar irradiance (e.g., in the morning and / or evening) and / or high temperatures, the DC voltage may no longer be sufficient to set a sufficiently high voltage at the inverter's AC input.

[0020] In the case of a DC storage unit, the DC voltage can depend, among other things, on the state of charge of the DC storage unit's battery. In particular, a fully charged battery can have a relatively high DC voltage, allowing the exchange of electrical power via the inverter to occur at a correspondingly high AC voltage. In the case of an electrolyzer, the power drawn by the electrolyzer can be adjusted based on the voltage, so that the electrolyzer, especially at high power levels, has a relatively high DC voltage and can draw electrical power from a correspondingly high AC voltage via the inverter.

[0021] Therefore, it is advantageous to vary the AC voltage in the AC system network depending on the DC voltage and to adjust the transformation ratio of the tap changer accordingly.

[0022] In some embodiments, the transformer includes a control unit for adjusting the tap changer. This control unit can operate autonomously, depending on the current voltage of the AC supply network and a (fixed) AC setpoint voltage for the power supply system. Additionally, the control unit can be influenced by the system controller. The control unit can then regulate the tap changer largely autonomously to approximate a predetermined AC setpoint voltage in the AC system network. To achieve this, the tap changer assumes different positions with varying turns ratios, depending on the AC supply network voltage. Furthermore, the control unit can receive external inputs, such as the AC setpoint voltage.Advantageously, the AC setpoint voltage can be variably specified via such an input and appropriately adjusted during the operation of the power supply system. The input, e.g., the AC setpoint voltage, is provided, for example, by the system controller, which is communicatively connected to the transformer or its control system.

[0023] In particular, a system controller can be used to carry out the procedure, which is communicatively connected to the inverter.

[0024] This method enables the optimization of energy supply systems, such as PV systems, storage systems, or consumer systems, especially large-scale systems in the megawatt range. Particularly in systems with central inverters that already have megawatt-scale output and are operated with a large number of PV strings, batteries, and / or consumers, the potential uses for the central inverter(s) can be expanded.

[0025] This method makes it possible to set the AC voltage as high as possible for a given DC voltage and reactive power. This is possible regardless of the direction of power flow; that is, it can be applied both to power supply systems that draw electrical power from the AC grid and to power supply systems that feed electrical power into the AC grid.

[0026] Furthermore, this method ensures that the DC voltage on the inverter's DC side is sufficiently high compared to the AC voltage on the inverter's AC side to enable regulated operation. The AC voltage on the inverter's AC side is only reduced when absolutely necessary, specifically when the inverter cannot generate a regulated AC voltage from a given DC voltage that would allow feed-in to the AC grid. This occurs, for example, with a relatively low MPP and / or battery voltage, or when a regulated DC voltage cannot be generated from a given AC voltage on the inverter's AC side, which is required for feed-in to the DC unit, such as a low-state-of-charge storage system or a low-power electrolyzer.

[0027] In one embodiment, the vector length is proportional to the modulation level of a pulse-width modulation (PWM) control system of the inverter. The vector length thus essentially represents the ratio between the DC voltage applied to the inverter and the amplitude of the AC voltage generated by the inverter for controlled power transfer. Depending on the specific control system used by the inverter, the vector length can exist in various equivalent forms and may be scaled.

[0028] The amplitude of the AC voltage vector generated by the inverter's control circuitry on the AC side can be expressed as a percentage of the maximum modulation of the current DC voltage. At 100%, the vector length corresponds exactly to the full scale of a sinusoidal signal within the given DC limits. Through so-called overmodulation, the vector length can increase to approximately 105% during operation without causing impermissible distortions and harmonic currents. Special clocking techniques can be used for this purpose, such as space vector modulation and / or modulation of the DC or AC center potential, or conventional modulation techniques can be combined with filter elements of the inverter to ensure that any distortions of the input currents remain within the permissible limits, even at vector lengths up to approximately 105%.

[0029] In one embodiment, the inverter follows the AC voltage on the AC system grid and applies the current to its AC side. In this embodiment, the inverter is not voltage-regulating, and the power supply system, while potentially providing grid support, is not, or at least not predominantly, operating as a grid-forming component in an island grid.

[0030] In one embodiment, the AC setpoint voltage is repeatedly determined as a function of the vector length, wherein the vector length depends in particular on the instantaneous voltage on the DC side of the inverter.

[0031] In one embodiment, the vector length depends on the instantaneous apparent power exchanged with the AC supply network, in particular on the instantaneous active and / or reactive power of the inverters.

[0032] This method ensures that even with minimal DC voltage, such as the minimum MPP voltage of the PV system, a low state of charge in the storage system's battery, or low electrolyzer power, a vector length of 105% is not exceeded. For this purpose, the transformer's turns ratio can be dynamically and flexibly selected so that, with DC voltage present and reactive power required in overexcited operation, the AC voltage on the AC system network can be increased sufficiently to prevent the vector length from exceeding 105%.

[0033] The tap changer allows the voltage at the transformer's on-site winding to be adjusted to the AC setpoint voltage. This allows the AC setpoint voltage to be maintained by changing the turns ratio via the tap changer when the AC supply voltage fluctuates. Advantageously, the AC setpoint voltage is also adjusted to the current DC voltage during this process, thus utilizing any available control reserves regarding the vector length. This enables the transformer's turns ratio to be adapted not only to a changing AC supply voltage but also to a changing AC voltage on-site, for example, an AC voltage in the system increased by reactive power input.This flexible adaptation to both changing DC voltage and changing system-side AC voltage makes the operation of the system more flexible, robust and / or efficient.

[0034] The method makes it possible to select the highest possible AC voltage in the power supply system on the one hand, and on the other hand to ensure that the maximum vector length for trouble-free, controlled power exchange is not exceeded.

[0035] In a method for operating a power supply system connected to an AC supply network via a transformer, which exchanges electrical power with the AC supply network via the transformer, wherein the transformer is connected on one side to the AC supply network and on the other side to an AC system network of the power supply system, the power supply system has more than one inverter, which exchange electrical power between a DC unit on the DC side of the inverter and the AC system network on the AC side of the inverter. The method comprises the following steps: • Receiving a vector length from each inverter by the system controller, wherein the respective vector length is determined in the respective inverter and includes the respective ratio of the amplitudes of the voltages on the AC side and on the DC side of the respective inverter, • Setting an AC setpoint voltage for the AC system network by the system controller as a function of a vector length derived from the respective vector lengths, which in particular includes a vector length averaged from the respective vector lengths, wherein the AC setpoint voltage is reduced if the derived vector length exceeds a first predefinable limit, and wherein the AC setpoint voltage is increased if the derived vector length falls below a second predefinable limit, wherein the second limit is below the first limit. • Transmitting the AC setpoint voltage to the transformer, wherein a tap changer of the transformer is arranged to set a turns ratio such that the product of the voltage in the AC supply network and the turns ratio approximates the AC setpoint voltage, wherein the distance between the first and second limit values ​​depends on the tap width of the tap changer and is in particular proportional to the relative voltage change of at least 3 switching stages of the tap changer.

[0036] For this purpose, the system controller can, for example, detect the vector lengths of the modulation of the respective inverters. This can be achieved, for example, by the inverters communicating their respective current vector lengths (or equivalent values) of their current control to the system controller. An average value can then be calculated from the vector lengths of several or all inverters, measured as a percentage, that are connected to a common system network with a common transformer and tap changer. In one embodiment of the method, the averaged vector length comprises an arithmetic mean of the respective vector lengths of the individual inverters.

[0037] In one embodiment of the method, the derived vector length is determined from the respective low-pass filtered vector lengths of the individual inverters. For example, the vector lengths and / or the mean derived therefrom, e.g., the arithmetic mean, can be further low-pass filtered with a time constant to obtain a moving average of the vector lengths of all inverters. Preferably, the time constant is large relative to the control time, e.g., longer than one minute. The derived vector length can thus depend on a moving average of the respective vector lengths.

[0038] In one embodiment, the variation of the individual vector lengths around the mean value, e.g., the moving average, can be determined. This is particularly advantageous for determining a maximum value of the current vector lengths. This variation can be large, for example, with DC units of different sizes or operating differently in the power supply system. It can also be relatively small within a system, especially if the inverters connected to a common transformer are homogeneously constructed, i.e., they have comparable DC units, are exposed to comparable environmental conditions, and are otherwise operated largely symmetrically. This can apply to PV systems, storage systems, and large consumers such as electrolyzers.

[0039] In one embodiment of the method, the AC setpoint voltage is reduced to decrease the vector lengths when the derived vector length of the inverters exceeds a first predefinable limit, wherein the first limit is between 95% and 105%, preferably between 98% and 102%, and most preferably at 99%. If the derived vector length continues to increase, the AC setpoint voltage is further reduced until a vector length of approximately the first limit is reached again, i.e., e.g., 99% or less. In a power supply system with one inverter, the vector length of the single inverter can be used instead of the derived vector length.

[0040] Since the vector length can generally exceed 99% during inverter operation, but should not exceed 105%, the range between 99% and 105% represents a control reserve. This reserve should be large because the control of the tap changer is comparatively slow, and other boundary conditions can change rapidly between control interventions. For example, the reactive power setpoint for the power supply system could suddenly be adjusted by the grid operator. If the inverters generate the corresponding reactive power, this could lead to a change in the AC voltage within the power supply system and thus to a change in the vector length for a given DC voltage.

[0041] With a derived vector length – or a vector length for a single inverter – between 97% and 99%, a change in the AC setpoint voltage might be avoided. However, in one embodiment, a change in the AC setpoint voltage can occur if the individual vector lengths exhibit a large variation, i.e., in particular if the maximum value of the current vector lengths is above the first limit value plus a tolerance band, where the tolerance band lies within the aforementioned control reserve.

[0042] If the inverter vector length is unavailable or not directly available for transmission to the system controller, an alternative is to measure the effective voltage directly on the AC side of the inverter – before or after any smoothing reactor – and compare it to the DC voltage on the DC side of the inverter. The following relationships can be determined from such a voltage measurement: • Vector length in % = 100 x Effective line voltage directly at the inverter bridge output x √2 / U_DC • Vector length in % = 100 x Effective terminal voltage at the AC outputs of the inverter. xkx √2 / U_DCU_DC is the DC voltage and k is a correction factor that depends on the current through the sine wave filter choke. To calculate k, the current, the phase angle between current and voltage, the inductance of the sine wave filter choke, and the capacitance of the sine wave filter capacitor can be used.

[0043] In one embodiment of the method, the AC setpoint voltage is increased when the vector length V and / or the derived vector length reaches a second predefinable limit, wherein the second limit is between 90% and 100%, preferably between 96% and 98%. Preferably, the second limit is below the first limit.

[0044] Once the AC setpoint voltage has been reduced to achieve a derived vector length—or a vector length for a single inverter—of less than the first limit value, e.g., 99%, the AC setpoint voltage will only be increased again when the derived vector length falls below a second limit value. This second limit value can be, for example, between 90% and 100%, preferably between 96% and 98%, e.g., at approximately 97%. The smaller the derived vector length becomes, the further the AC setpoint voltage is increased to achieve a derived vector length greater than the second limit value. The same applies to the vector length in power supply systems with a single inverter. Once the AC setpoint voltage has reached a preset maximum voltage and / or the highest stage of the tap changer has been reached, no further increase occurs.

[0045] In this method, the distance between the first and second limit values ​​depends on the step width of the tap changer and is in particular proportional to the relative voltage change of at least 3 switching steps of the tap changer.

[0046] This is one possible definition of a tolerance range within which the vector length can move without the tap changer undergoing any switching operations.

[0047] The difference between the first and second threshold values ​​corresponds to a hysteresis width. The hysteresis width can preferably be large enough to encompass the voltage change of at least three switching stages; with a typical "difference" between the switching stages of, for example, 0.65% of the nominal voltage, the hysteresis width could therefore be, for example, at least 2% of the vector length. This results in a tolerance range within which the vector length can vary without triggering any switching operations of the tap changer.

[0048] In one embodiment of the method, the maximum number of changes to the AC setpoint voltage per unit of time can be specified and is, for example, between 30 and 70 times per day, preferably between 40 and 60 times per day.

[0049] The process should be configured so that the average daily AC setpoint voltage is not changed more than a predefined number of times, e.g., 50 times, triggering a switching operation in the tap changer. An adaptive method can be used to count the switching operations per day. If it is anticipated that more than the predefined number of changes to the AC setpoint voltage will be necessary (e.g., more than 50), the time constant for calculating the moving average of the vector length can be increased. This time constant can be limited, e.g., to a maximum of 20 minutes. If fewer than the predefined number of changes to the AC setpoint voltage are recorded on a given day, the time constant for calculating the moving average can be reduced accordingly. Such an adaptive method can, in particular, start with a time constant in the middle range, i.e., around half the limit, e.g., 10 minutes.

[0050] Another way to reduce the number of cycles is, for example, to increase the hysteresis width. In the example described above, the hysteresis is 2%. However, it can also be increased, for example to 3% or more, if more than the predefined number of changes in the AC setpoint voltage per day are registered.

[0051] The rate of change of the AC setpoint voltage can be limited to a speed that depends on the nominal voltage of the power supply system. For example, the rate of change of the AC setpoint voltage can be limited to a maximum of Un per unit of time, where Un is the nominal voltage of the system and a suitable unit of time is, for example, 1 hour. Intervals between two switching operations on the tap changer of less than 1 minute should preferably be avoided.

[0052] Another way to reduce the number of cycles, i.e., the number of changes in the AC setpoint voltage and thus the number of switching operations on the tap changer, is, for example, to limit the rate of change of the AC setpoint voltage more strongly, for example by increasing the time constant in the above example to, for example, 2 hours.

[0053] In one embodiment of the method, the specified AC setpoint voltage is above a predefinable minimum value and / or below a predefinable maximum value, wherein the minimum value and / or the maximum value depends on a nominal voltage for the AC system network.

[0054] The AC setpoint voltage preferably does not fall below an adjustable minimum value. The AC setpoint voltage preferably does not exceed an adjustable maximum value. The minimum and maximum values ​​typically lie within a tolerance band for the system's internal mains voltage. Example: • Maximum system mains voltage = 110% → maximum AC setpoint voltage = 107% • Minimum system mains voltage = 90% → minimum AC setpoint voltage = 92%

[0055] These percentage values ​​refer to the system's nominal voltage. An asymmetry in the tolerance band and / or the minimum and maximum values ​​relative to the system's nominal voltage is possible.

[0056] Another method serves to operate a transformer in a power supply system, which is connected to an AC supply network via the transformer. Electrical power is exchanged between the power supply system and the AC supply network via the transformer, with the transformer being connected to the AC supply network on one side and to an AC system network of the power supply system on the other. The power supply system has at least one inverter, which exchanges electrical power between a DC unit on the DC side of the inverter and the AC system network on the AC side of the inverter. The power supply system also has a system controller, and the transformer has a control unit. The method is characterized by the following steps: • Receiving an AC setpoint voltage for the AC plant network by the transformer's control unit from the plant controller of the power supply system, • Adjusting the transformer's turns ratio using a tap changer so that the product of the voltage in the AC supply network and the turns ratio equals the AC setpoint voltage, • where the AC target voltage was determined using at least one parameter of the power conversion of the at least one inverter.

[0057] If the energy supply system is configured as a PV system, for example, with the DC units comprising PV generators, the positions of the tap changer can follow a typical daily cycle of the PV generators' MPP voltage. For instance, the PV system starts in the morning with a low DC voltage and a correspondingly low AC voltage to enable the earliest possible start-up. The MPP voltage, and thus the DC voltage at the inverter, rises soon after sunrise, allowing the AC voltage to be increased to minimize losses. At high temperatures, particularly around midday, the MPP voltage can drop again, potentially requiring a reduction in the AC voltage, which then rises again later, reaching a low value at sunset, analogous to the sunrise level.

[0058] If the power supply system is designed as an energy storage system, for example, by including batteries in the DC units, the positions of the tap switch can, for example, follow the state of charge of the batteries.

[0059] For example, when the state of charge is low and the DC voltage is correspondingly low, charging or discharging can be performed with the lowest possible AC voltage to utilize the battery's voltage range as fully as possible. As the state of charge increases, the DC voltage also rises, allowing the AC voltage to be increased to minimize losses.

[0060] If the power supply system is designed for large consumers, for example, by including electrolyzers in the DC units, the tap changer positions can be adjusted to reflect the electrolyzer's target power, which can be set using the DC voltage. For instance, at a low target power and correspondingly low DC voltage, the inverter can supply the electrolyzer from the lowest possible AC voltage to maximize the electrolyzer's power range. A high target power for the electrolyzer requires a correspondingly high DC voltage, which the inverter can also generate from a high AC voltage, thus increasing the AC voltage, particularly to minimize losses.

[0061] In addition to the characteristics of the DC unit, the operation of the inverter itself can influence the AC voltage in the power supply system. In particular, reactive power generated by the inverter can lead to a change in the AC voltage due to the inductive properties of the power supply system's installations. The method according to the invention can take these influences into account by making the power conversion parameter dependent on the AC voltage at the inverter's output, and thus, for example, enabling operation with voltage-increasing reactive power in the overexcited range by reducing the AC voltage based on a suitably lowered AC setpoint voltage.

[0062] The transformer's tap changer adjusts the turns ratio based on the AC setpoint voltage and the AC voltage in the AC supply network. Given a specific AC setpoint voltage, the turns ratio changes when the voltage in the AC supply network changes. Conversely, given a specific voltage in the AC supply network, the turns ratio changes when the AC setpoint voltage changes. Therefore, transmitting the AC setpoint voltage to the transformer allows the tap changer to adjust the turns ratio and thus optimally adapt the AC voltage in the power supply system. The transformer is, for example, a high-voltage transformer, such as a 120 MVA transformer.

[0063] The transformer of the power supply system, located between an inverter and a higher-level AC supply network, contains a tap changer that can change the transformer's turns ratio into, for example, 31 steps. This allows the system voltage to be varied within a range of, for example, + / -10% (0.65% per step) for a given AC voltage in the supply network. Larger ranges of variation up to + / -20% are also possible. The midpoint of the range can correspond to a nominal voltage, allowing for symmetrical changes of the same amount, or it can deviate from a nominal voltage, resulting in an asymmetrical range around the nominal voltage (e.g., +15%, -5%).

[0064] In one embodiment, the transformer and / or its tap changer or its control unit can be controlled communicatively, in particular via a fieldbus, and supplied with the AC set voltage in this way.

[0065] The control of the transformer for the tap changer can be used to make the use of high-power central inverters in the megawatt range particularly efficient at high AC voltages, which may be influenced by reactive power injection by the inverter itself, and at the same time at at least temporarily low DC voltages.

[0066] In one embodiment of the method, the at least one parameter comprises a vector length, such that a change in the voltage on the DC side and / or a change in the apparent power supplied by the at least one inverter causes a change in the transformation ratio of the transformer.

[0067] In one embodiment of the method, a change in the reactive power of at least one inverter causes a change in the transformer's turns ratio. This is preferably achieved by changing the AC setpoint voltage in response to a change in the inverter's reactive power or the resulting change in vector length.

[0068] The vector length is by definition exactly 100% if the phasor voltage of the inverter, i.e., the magnitude of the rotating phase vector of the AC voltage on the AC side of the inverter, exactly matches the set DC voltage, especially if no reactive power is generated by the inverter.

[0069] A key advantage of using the vector length as a parameter or as part of a parameter is that it eliminates the need for the complex analysis of all possible individual influencing factors. The vector length is a particularly advantageous value because it is generally readily available and inherently encompasses most, if not all, of the relevant influencing factors for the optimal control of the tap changer.

[0070] In one embodiment, as an alternative or additional to transmitting the AC setpoint voltage to the tap changer's control unit, the tap changer can also be accessed directly, so that, instead of the AC setpoint voltage, the turns ratio itself, for example, can be used as the control variable. Then, the system controller could, as an alternative or additional to controlling the inverters, also control the tap changer. In particular, the system controller could then also react to any changes in the mains voltage in the upstream AC supply network, whereby the mains voltage in the AC supply network could be made available, for example, by an AC current measurement at the grid connection point. The system controller could then also, as an alternative or additional to controlling the transformer, implement safety functions, specific switching sequences, delays, etc.This allows the system controller to also take over at least some of the functions of a largely self-sufficient control unit for the transformer.

[0071] A power supply system is connected to an AC supply network via a transformer. The transformer is connected to the AC supply network on one side and to the AC system network of the power supply system on the other. The power supply system includes an inverter, which is connected to a DC unit on one side and to the AC system network on the other. A suitable system controller for such a power supply system is installed. • to receive a vector length of the inverter, wherein the vector length is determined in the inverter and includes the ratio of the amplitudes of the voltages on the AC side and on the DC side of the inverter, • to determine an AC setpoint voltage for the AC system network based on the vector length, wherein the AC setpoint voltage is reduced if the vector length exceeds a first predefinable limit, and wherein the AC setpoint voltage is increased if the vector length falls below a second predefinable limit, the second limit being below the first limit, • to transmit the determined AC setpoint voltage to the transformer in order to adjust the transformer's turns ratio using a tap changer so that the product of the voltage in the AC supply network and the turns ratio approximates the AC setpoint voltage, • wherein the distance between the first and second limit depends on the step width of the tap changer and is in particular proportional to the relative voltage change of at least 3 switching steps of the tap changer.

[0072] A sophisticated system controller for a corresponding energy supply system with at least two inverters receives vector lengths from the inverters and determines the AC setpoint voltage analogously based on a vector length derived from the respective vector lengths.

[0073] Preferably, the AC setpoint voltage for the AC system network is determined in such a way that the highest possible AC voltage can be achieved on the AC system network during operation of the system.

[0074] A power supply system compliant with the requirements comprises a transformer, an AC power grid, at least one inverter and at least one DC unit, wherein the power supply system includes a controller configured to carry out one of the above methods. Brief description of the characters

[0075] The following section provides a more detailed explanation of exemplary applications using figures. Fig. Figure 1 schematically shows an embodiment of a method for operating an energy supply plant; Fig. Figure 2 schematically shows an energy supply system; Fig. Figure 3 shows exemplary time profiles of various quantities in an energy supply system. Character description

[0076] In Fig. 1 is an embodiment of a method for operating an energy supply plant (10, see Fig. 2) shown. The power supply system 10 comprises a transformer 14 with a tap changer 16, an AC power grid 18, and two inverters 22 and 24. The power supply system 10 also comprises a controller 20, which is communicatively connected, e.g., via a fieldbus 30, to the inverters 22 and 24 and to the transformer 14 or to the tap changer 16 of the transformer 14 (or its control unit 16, not shown). The power supply system 10 is connected to an AC power grid 12 via the transformer 14.

[0077] In step S1, the system controller 20 of the power supply system 10 receives a vector length V from each of the respective inverters 22 and 24. In step S2, the system controller 20 sets an AC setpoint voltage for the AC system network 18 as a function of a vector length derived from the vector lengths V. In step S3, the system controller 20 transmits the AC setpoint voltage to the transformer 14, whereby the tap changer 16 of the transformer 14 is configured to set a turns ratio T such that the product of the voltage in the AC supply network 12 and the turns ratio T yields the AC setpoint voltage. In step S4, the tap changer 16 adjusts the turns ratio T accordingly.

[0078] In this process, the plant controller 20 transmits an AC setpoint voltage to the tap changer 16 in step S3. The actual control signals, i.e., the setting or changing of the transformation ratio of the transformer 14, are carried out by the tap changer 16 or by a control unit of the transformer 14 that controls the tap changer 16.

[0079] Fig. Figure 2 shows an embodiment of the energy supply system 10 with the system controller 20, which is communicatively connected to the inverters 22, 24 via a wired or wireless communication structure, e.g. via the fieldbus 30.

[0080] The power supply system 10 exchanges electrical power with the AC supply network 12 via transformer 14. Transformer 14 is connected to the AC supply network 12 on one side and to the AC system network 18 of the power supply system 10 on the other. The turns ratio of transformer 14 can be adjusted via tap changer 16. Transformer 14 also includes a control unit for operating tap changer 16. The power supply system 10 has two inverters 22 and 24, which exchange electrical power between a PV string 26 and 28 on the DC side of inverters 22 and 24, respectively, and the AC system network 18 on the AC side of inverters 22 and 24.

[0081] The PV strings 26, 28 shown are examples of DC units in a power supply system 10 and can easily be replaced by storage devices, e.g., batteries, or by loads, e.g., electrolyzers. No significant changes are required for this, neither to the inverters 22, 24 nor to the procedure described above.

[0082] The control unit of transformer 14 is configured to receive the AC setpoint voltage for the AC plant network 18 from the plant controller 20 of the power supply plant 10. The control unit of transformer 14 is further configured to adjust the turns ratio of transformer 14 by means of the tap changer 16 so that the product of the voltage in the AC supply network 12 and the turns ratio yields the AC setpoint voltage.

[0083] If the specified AC setpoint voltage changes significantly and / or the voltage in the AC supply network 12 changes sharply, the tap changer 16 ensures that several steps are switched in quick succession to reach the step at which the voltage applied to the transformer 14 on the system side corresponds to the AC setpoint voltage. The system-side voltage at the transformer 14 does not need to be explicitly measured or determined; rather, the tap changer 16 can be set so that the product of the (measured) voltage in the AC supply network 12 and the turns ratio T yields the AC setpoint voltage. This compensates for changes in the voltage in the AC supply network 12; however, it neglects the fact that this product is usually not equal to the system-side voltage at the transformer 14 and certainly not equal to the voltages at remote points within the power supply system 10.Such deviations occur in particular due to resistive and inductive effects at high active and reactive power levels in the power supply system 10 and are taken into account in the procedure by setting the transformation ratio T using the AC setpoint voltage, wherein the AC setpoint voltage is determined by the system controller 20 using the vector length derived from the respective vector lengths V of the inverters 22, 24.

[0084] If, for example, the derived vector length lies outside a desired range, particularly if it reaches an upper limit, new setpoints for the tap changer 16 are generated in the transformer 14. If these setpoints cannot be implemented by the tap changer 16, or can only be implemented with a delay, e.g., due to overheating or a defect in the tap changer 16, a corresponding feedback signal is generated by the tap changer 16 and transmitted to the plant controller 20. The plant controller 20 can then, if necessary, adjust the active and reactive power setpoints for the inverter(s) 22, 24 to modify the vector length V in another way, in particular to reduce it and limit it to a maximum vector length of, for example, 99%.

[0085] The inverter 22, 24 features a bridge circuit in which AC-side alternating current (or AC-side alternating voltage) is converted into DC-side direct current (or DC-side direct voltage), or vice versa, primarily by pulsed control of semiconductor switches. The inverter 22, 24 provides a "chopped" DC voltage to the AC side of the bridge circuit, which is applied to an inductor, for example, a choke in an output filter of the inverter 22, 24. On the other side, i.e., "downstream" the choke, the AC voltage of the AC system network is applied. This results in a standard sinusoidal AC current that follows the AC voltage, provided that the instantaneous average value of the chopped DC voltage is always higher than the instantaneous AC voltage downstream of the choke. The inverter 22, 24 thus provides an AC voltage to its bridge circuit as a pulse-width modulated DC voltage.To ensure that this functions correctly, particularly with inverters featuring DC-side intermediate circuits and conventional bridge circuits, allowing for the exchange of adjustable active power between the DC unit 26, 28 and the AC system network 18, the DC-side voltage should be at least as high as the peak value of the AC network voltage of the AC system network 18. This corresponds to a vector length V of 100%. To generate reactive power, the DC voltage may even need to be higher relative to the peak value of the AC voltage on the AC system network 18 due to the phase shift between current and voltage. By appropriately designing the choke and / or using a special modulation method, an inverter 22, 24 can also be operated with vector lengths V greater than 100%, especially in three-phase networks. The operating range with vector lengths V greater than 100% is used here as a so-called control reserve.

[0086] The current vector length V during the operation of an inverter 22, 24 is particularly influenced by: - the current DC voltage on its DC side, which is variable, - the current apparent power in the energy supply system 10, divided into active power and reactive power, both of which are variable, - the number of inverters 22, 24 on the same transformer 14 and the impedances between inverters 22, 24 and the grid connection point to the AC supply network 12; these factors are usually plant-specific and therefore largely constant, provided that all inverters 22, 24 are in regular operation.

[0087] Reactive power influences the vector length V both directly at the respective inverters 22, 24 (higher reactive power = higher current = longer phase vector = greater vector length V), and indirectly through its influence on the voltages along the installations in the power supply system 10. The resulting inductive decoupling impedance between the terminals of the bridge circuit of inverters 22, 24 and the grid connection point includes an AC choke in inverters 22, 24, the inductance of transformer 14 with tap changer 16, the inductances of the extended AC lines of the AC system network 18, and, for example, other intermediate transformers. Particularly in overexcited operation, reactive power causes a voltage increase and, especially due to the inductance of transformer 14, an increase in the system-side AC voltage at transformer 14.

[0088] In Fig. Figure 3 shows time profiles for a vector length V, a DC voltage U_DC on the DC side of an inverter 22, 24, a reactive power Q generated by the inverter and the turns ratio T of the transformer 14.

[0089] The vector length V can represent the vector length of one of the inverters 22, 24, or the vector length derived from the vector lengths of inverters 22, 24, provided that the inverters 22, 24 are operated largely equivalently. The DC voltage U_DC can represent the DC voltage of one of the DC units 26, 28, or an average of the DC voltages of the DC units 26, 28, provided that the DC units 26, 28 are largely homogeneous. The reactive power Q can represent a setpoint that is configured at the grid connection point of the power generation plant. The turns ratio T is inversely proportional to the resulting AC setpoint voltage, such that a low AC setpoint voltage, given a constant voltage in the AC supply network 12, requires a high turns ratio T, and vice versa.

[0090] At time t1, the voltage U_DC is too low, so neither inverter 22 nor 24 is yet operational. The vector length V is therefore shown as a dashed line, since it can only be estimated from the DC and AC voltages, but is not yet available as a parameter due to the lack of active control. The transformation ratio T is set to the maximum possible value n. max This ensures that the AC voltage in the system network 18 is as low as possible, in order to enable the inverters 22, 24 to start up at the lowest possible DC voltage U_DC.

[0091] At time t2, the DC voltage has reached a value sufficient to operate inverters 22 and 24. Due to the high transformation ratio T=n maxAt time t2, the vector length falls below the lower limit of 97%. The system controller 20 can then increase the AC setpoint voltage, causing the tap changer 16 to reduce the turns ratio T of the transformer 14. Due to the continuing increase in DC voltage U_DC, this process repeats twice more until time t3.

[0092] Between t3 and t4, a relatively low turns ratio T is set, meaning the AC voltage in the plant network 18 is relatively high. The DC voltage U_DC exhibits slight fluctuations, but these are so small that the vector length V remains between the lower limit at 97% and the upper limit at 99%.

[0093] At time t4, the DC voltage U_DC has dropped to such an extent that the vector length V exceeds the upper limit at 99%. Consequently, the system controller 20 lowers the AC setpoint voltage, causing the tap changer 16 to increase the turns ratio T of the transformer 14. This process repeats shortly after t4 due to the slightly further decrease in the DC voltage U_DC.

[0094] At time t5, the reactive power Q changes abruptly. This can be triggered, for example, by the implementation of a mandatory reactive power requirement by a grid operator or by a reaction of the power generation plant 10 to a grid fault in the AC supply network 12. In the example according to Fig. 3. The generation of reactive power by inverters 22, 24 leads to a reduction in the AC voltage in the system network 18, whereupon the vector length V also decreases and falls below the lower limit of 97%. Consequently, the AC setpoint voltage can be increased and the turns ratio T can be reduced accordingly; this occurs in Fig. 3. With a certain time delay, the number of switching operations can be limited, for example, by waiting for the control system to settle and thus avoiding unnecessary switching operations due to temporary effects. The time delay until the change in the transmission ratio T is also uncritical, since a short vector length only causes unnecessary losses. After the change in the transmission ratio T, the vector length V initially remains constant at approximately 99%.

[0095] At time t6, the reactive power Q changes in the opposite direction, which in this example leads to an increase in the AC voltage in the system network 18. Accordingly, the vector length V also increases and initially reaches a value above the upper limit of 99%. This is uncritical to a certain extent and for a certain period of time, as the inverters 22, 24 can generally be operated with vector lengths up to approximately 105% within a certain control reserve. However, at time t7, the vector length V increases significantly again due to another change in the reactive power Q, so that the system controller 20 reacts immediately and significantly reduces the AC setpoint voltage, thus increasing the turns ratio T accordingly in order to keep the vector length V below the upper limit of 99%.

[0096] At time t8, the reactive power is set to zero, causing the AC voltage in the plant network to drop again and the vector length to fall below the lower limit of 97%. Subsequently, the turns ratio T is also reduced, possibly with a certain time delay, to increase the AC voltage within the permissible range and thereby, in particular, to reduce losses.

[0097] At time t9, the DC voltage U_DC continuously decreases and reaches such a low value at time t10 that inverters 22 and 24 can no longer be operated and shut down (the reverse of the switch-on at time t2). Due to the drop in DC voltage between t9 and t10, the vector length V continuously increases and is repeatedly reduced below the upper limit of 99% by increasing the turns ratio T, until the maximum turns ratio n is reached. maxhas been reached and further increases in vector length can no longer be counteracted.

[0098] The in Fig. The three time series shown are for illustrative purposes only and are not to scale. In principle, such or similar series can occur in energy supply systems that include DC units in the form of PV generators, energy storage systems, and / or consumers. In particular, Fig. Figure 3 illustrates a typical daily cycle of a PV system, where the MPP voltage - and thus the DC voltage - has a steep rise at sunrise and a - somewhat less - steep drop at sunset and is largely constant in between, but can still vary due to varying temperatures, for example.

[0099] Specifically, the time sequences in the period t1 to t3 can illustrate, for example, the behavior of a PV system at sunrise or the charging and starting of an electrolyzer. The behavior in the period t3 to t5 can occur particularly in a PV system exposed to fluctuating temperatures. In the period t5 to t8, the behavior is mainly influenced by external factors, especially reactive power specifications, and can therefore occur in energy supply systems with all types of DC units. The time sequences in the period t9 to t10 are particularly typical for PV systems at sunset or for energy storage systems at high discharge power.

[0100] This method ensures that the AC voltage is kept as high as possible and only reduced when absolutely necessary, particularly when the inverter cannot generate a regulated AC voltage from a given DC voltage that allows feed-in from the DC unit (PV, storage, etc.) into the AC grid, or conversely, when a regulated DC voltage cannot be generated from a given AC voltage that allows feed-in into the DC load (storage, electrolysis, etc.). Reference symbol list 10 PV systems 12 AC supply network 14 Transformer 16-step switch 18 AC plant network 20 system controllers 22 inverters 24 inverters 26 PV string 28 PV string 30 Data bus S1, S2, S3, S4 Procedure steps Q Apparent power T translation ratio U_DC voltage on the DC side V vector length t1 - t10 Time points

Claims

[1] A method for operating a power supply system (10) which is connected to an AC supply network (12) via a transformer (14) and exchanges electrical power with the AC supply network (12) via the transformer (14), wherein the transformer (14) is connected on a first side to the AC supply network (12) and on a second side to an AC system network (18) of the power supply system (10), wherein the power supply system (10) has at least one inverter (22, 24) which exchanges electrical power between a DC unit (26, 28) on the DC side of the inverter (22, 24) and the AC system network (18) on the AC side of the inverter (22, 24), comprising the steps: • Receiving a vector length (V) of the at least one inverter (22, 24) by a system controller (20), wherein the vector length (V) is determined in the at least one inverter (22, 24) and comprises the ratio of the amplitudes of the voltages on the AC side and on the DC side of the at least one inverter (22, 24), • Determining an AC target voltage for the AC system network (18) by the system controller (20) as a function of the vector length (V), wherein the AC target voltage is reduced if the vector length (V) exceeds a first predeterminable limit value, and wherein the AC target voltage is increased if the vector length (V) falls below a second predeterminable limit value, wherein the second limit value is below the first limit value, • Transmitting the AC target voltage to the transformer (14), wherein a tap changer (16) of the transformer (14) is configured to set a transformation ratio (T) such that the product of the voltage in the AC supply network (12) and the transformation ratio approximates the AC target voltage, • wherein the distance between the first and second limit values ​​depends on the step width of the step switch (16) and is in particular proportional to the relative voltage change of at least 3 switching steps of the step switch. [2] Method according to claim 1, characterized by that the vector length (V) is proportional to the control level of a PWM control of the at least one inverter (22, 24). [3] Method according to claim 1 or 2, characterized bythat the AC target voltage is repeatedly determined as a function of the vector length (V), wherein the vector length (V) depends in particular on the instantaneous voltage (U_DC) on the DC side of the at least one inverter. [4] Method according to claim 3, characterized by that the vector length (V) depends on the instantaneous apparent power (Q) exchanged with the AC supply network, in particular on the active and / or reactive power of the at least one inverter (22, 24). [5] A method for operating a power supply system (10) which is connected to an AC supply network (12) via a transformer (14) and exchanges electrical power with the AC supply network (12) via the transformer (14), wherein the transformer (14) is connected on a first side to the AC supply network (12) and on a second side to an AC system network (18) of the power supply system (10), wherein the power supply system (10) has more than one inverter (22, 24) which exchange electrical power between a respective DC unit (26, 28) on the DC side of the inverter (22, 24) and the AC system network (18) on the AC side of the inverter (22, 24), comprising the steps: - receiving a respective vector length (V) of the inverters (22, 24) by a system controller (20), wherein the respective vector length (V) is determined in the respective inverter (22, 24) and comprises the respective ratio of the amplitudes of the voltages on the AC side and on the DC side of the inverter (22, 24), - setting an AC target voltage for the AC system network (18) by the system controller (20) as a function of a vector length derived from the respective vector lengths (V), which vector length in particular comprises a vector length averaged from the respective vector lengths (V), wherein the AC target voltage is reduced if the derived vector length (V) exceeds a first predeterminable limit value, and wherein the AC target voltage is increased if the derived vector length (V) falls below a second predeterminable limit value, wherein the second limit value is below the first limit value, - transmitting the AC target voltage to the transformer (14), wherein a tap changer (16) of the transformer (14) is configured to set a transformation ratio (T) such that the product of the voltage in the AC supply network (12) and the transformation ratio approximates the AC target voltage, wherein the distance between the first and second limit values ​​depends on the step width of the tap changer (16) and is in particular proportional to the relative voltage change of at least 3 switching steps of the tap changer. [6] Method according to claim 5, characterized by that the derived vector length is determined from the respective low-pass filtered vector lengths (V) of the respective inverters (22, 24). [7] Method according to claim 5 or 6, characterized by that the averaged vector length comprises an arithmetic mean of the respective vector lengths (V) of the respective inverters (22, 24). [8] Method according to one of claims 1 to 7, characterized by that the first limit value is between 95% and 105%, preferably between 98% and 102%. [9] Method according to one of claims 1 to 8, characterized by that the second limit value is between 90% and 100%, preferably between 96% and 98%. [10] Method according to one of the preceding claims, characterized by that the maximum number of changes in the AC setpoint voltage per unit of time can be specified and is, for example, between 30 and 70 times per day, preferably between 40 and 60 times per day. [11] Method according to one of the preceding claims, characterized by that the specified AC target voltage is above a predeterminable minimum value and / or below a predeterminable maximum value, wherein the minimum value and / or the maximum value depends on a nominal voltage for the AC system network (18). [12] System controller (20) for a power supply system (10) which is connected to an AC supply network (12) via a transformer (14), wherein the transformer (14) is connected on a first side to the AC supply network (12) and on a second side to an AC system network (18) of the power supply system (10), wherein the power supply system (10) has an inverter (22, 24) which is connected on a DC side to a DC unit (26, 28) and on an AC side to the AC system network (18), wherein the system controller (20) is configured • to receive a vector length (V) of the inverter (22, 24), wherein the vector length (V) is determined in the inverter (22, 24) and comprises the ratio of the amplitudes of the voltages on the AC side and on the DC side of the inverter (22, 24), • to determine an AC target voltage for the AC system network (18) based on the vector length (V), wherein the AC target voltage is reduced if the vector length (V) exceeds a first predeterminable limit value, and wherein the AC target voltage is increased if the vector length (V) falls below a second predeterminable limit value, wherein the second limit value is below the first limit value, • to transmit the determined AC target voltage to the transformer (14) in order to set a transformation ratio (T) of the transformer (14) by means of a tap changer (16) such that the product of the voltage in the AC supply network (12) and the transformation ratio (T) approximates the AC target voltage, wherein the distance between the first and second limit values ​​depends on the step width of the step switch (16) and is in particular proportional to the relative voltage change of at least 3 switching steps of the step switch. [13] System controller (20) for a power supply system (10) which is connected to an AC supply network (12) via a transformer (14), wherein the transformer (14) is connected on a first side to the AC supply network (12) and on a second side to an AC system network (18) of the power supply system (10), wherein the power supply system (10) has at least two inverters (22, 24), each of which is connected on a DC side to a DC unit (26, 28) and on an AC side to the AC system network (18), wherein the system controller (20) is configured • to receive vector lengths (V) of the inverters (22, 24), wherein the respective vector length (V) is determined in the respective inverter (22, 24) and comprises the respective ratio of the amplitudes of the voltages on the AC side and on the DC side of the inverter (22, 24), • to determine an AC target voltage for the AC system network (18) based on a vector length derived from the respective vector lengths (V), which in particular comprises a vector length averaged from the respective vector lengths (V), wherein the AC target voltage is reduced if the derived vector length exceeds a first predeterminable limit value, and wherein the AC target voltage is increased if the derived vector length falls below a second predeterminable limit value, wherein the second limit value is below the first limit value, • to transmit the determined AC target voltage to the transformer (14) in order to set a transformation ratio (T) of the transformer (14) by means of a tap changer (16) such that the product of the voltage in the AC supply network (12) and the transformation ratio (T) approximates the AC target voltage, wherein the distance between the first and second limit values ​​depends on the step width of the step switch (16) and is in particular proportional to the relative voltage change of at least 3 switching steps of the step switch. [14] Energy supply system (10) with a transformer (14), an AC system network (18), at least one inverter (22, 24) and at least one DC unit (26, 28), wherein the energy supply system (10) comprises a system controller (20) which is designed to carry out the method according to one of claims 1 to 4. [15] Energy supply system (10) with a transformer (14), an AC system network (18), at least two inverters (22, 24) and at least two DC units (26, 28), wherein the energy supply system (10) comprises a system controller (20) which is set up to carry out the method according to one of claims 5 to 11.

Images