METHOD FOR OPERATING AN INVERTER ARRANGEMENT AND INVERTER ARRANGEMENT FOR IMPLEMENTING THE METHOD

Description

[0001] The invention relates to a method for operating an inverter arrangement with at least one inverter for converting a direct voltage into an alternating voltage at a mains frequency, wherein in each inverter the direct voltage of a DC source connected to a DC input is transformed into an intermediate circuit voltage via a DC / DC converter and the intermediate circuit voltage is converted into the alternating voltage in a DC / AC converter with two power units, each connected to an AC output, which are mutually modulated by a control device of the respective inverter so that opposing voltages are output at the AC outputs of the power units, which alternating voltage corresponds to the sum of the opposing voltages at the AC outputs of the power units with respect to the negative pole of the intermediate circuit voltage, and the alternating voltage is fed into a supply network connected to the AC outputs.

[0002] Furthermore, the invention relates to an inverter arrangement with at least one inverter for converting a direct voltage into an alternating voltage at a mains frequency, wherein each inverter has a DC / DC converter for transforming the direct voltage of a DC source connected to a DC input into an intermediate circuit voltage, a DC / AC converter with two power units for converting the intermediate circuit voltage into the alternating voltage, wherein each inverter has a control device for modulating the power units of each inverter.

[0003] The term inverter array refers to arrays consisting of one or more inverters, each of which is connected to a corresponding DC source, particularly photovoltaic modules or strings of photovoltaic modules, at the input, and whose output is connected to the grid. In an inverter array with multiple inverters, all inverters can be controlled independently of one another or connected to each other via communication lines so that their functions can be coordinated.

[0004] In particular, the present invention relates to modern single-phase inverters without a transformer (transformerless inverters), which convert a direct voltage into an alternating voltage at a grid frequency. The direct voltage of a direct voltage source, for example a photovoltaic module, is transformed into an intermediate circuit voltage via a DC / DC converter, and the intermediate circuit voltage is converted in a DC / AC converter into the desired alternating voltage, which is used to operate consumers or to feed into a supply grid. The DC / AC converter has two power units with semiconductor switches, which are modulated accordingly by a control device. In contrast to inverters with transformers, transformerless inverters are significantly more efficient.

[0005] Inverters without transformers and without galvanic isolation generate a grid-dependent potential change through modulation (pulse width modulation PWM) at the DC source, particularly at the photovoltaic generator. This potential change generates leakage currents to earth due to unavoidable parasitic capacitances and insulation resistances between components of the inverter arrangement, particularly between the photovoltaic modules, and earth. Unacceptably high leakage currents can result in a mandatory residual current device (RCD) being triggered when the inverter is connected to the supply grid. If the RCD is triggered, other devices may also be unintentionally disconnected from the supply grid. Switching on the RCD requires manual intervention and therefore the presence of a person. For this reason, it is common practice to measure the leakage current orto determine the responsible leakage capacitance and, if necessary, the insulation resistance and, if necessary, to prevent the inverter from being connected to the supply grid if the maximum permissible leakage current is exceeded. Only after the causes of the high leakage current have been eliminated and the residual current device has been closed again is the alternating voltage generated in the inverter available again to supply consumers with electrical energy or to feed into the supply grid. Before this time, the converted energy cannot be used, which results in a poorer energy balance. The operator of the inverter system thus loses valuable feed-in time, feed-in power and feed-in energy and therefore money. The lost feed-in time, power and energy increases proportionally with the size of the inverter system. In high-performance systems, this can result in very high yield losses.

[0006] EP 3 465 897 B1 describes a method for operating a single transformerless inverter. If a limit value for the leakage current is exceeded, the inverter is controlled accordingly or operated using a special clocking method that can reduce the leakage current. While this allows the inverter to be switched on in a way that would otherwise have been impermissible, the resulting energy balance is still reduced, and the operator must accept corresponding yield losses. Furthermore, determining the leakage current involves a certain amount of effort.

[0007] CN 110323959 A describes a method for operating an inverter and such an inverter, wherein harmonics of the mains frequency are reduced with a filter circuit at the output of the inverter and leakage currents are reduced via an additional capacitor and an additional compensation circuit.

[0008] EP 1 229 629 A2 describes an inverter with a possibility of detecting leakage currents.

[0009] The object of the present invention is to create an above-mentioned method for operating an inverter arrangement and an inverter arrangement for implementing this method, which is intended to ensure that even with higher expected leakage currents, the inverter arrangement can be connected to the supply grid and that the converted energy can then be utilized for a better energy balance. The method and the inverter arrangement should be as simple and cost-effective to implement as possible and should not incur any significant additional costs for the operator. Disadvantages of previous methods or devices should be avoided or at least reduced.

[0010] The object of the invention is achieved in terms of the method in that, before feeding the alternating voltage into the supply network, the leakage capacitance at the DC input of each inverter to earth is determined, and it is determined which AC output is connected to which line of the supply network, and taking into account the determined leakage capacitance and a defined maximum leakage current as well as the determined connection situation at the AC output, the power units of each inverter are asymmetrically modulated by the control device in such a way that the amplitudes of the voltages at the AC outputs of the power units, which correspond to the alternating voltage to be fed into the supply network, are of different sizes, in that the modulation of the power unit of the DC / AC converter connected to the phase connection of the supply network is carried out in such a way that the voltage with the greater amplitude results at the AC output of this power unit,and the modulation of the power section of the DC / AC converter connected to the neutral conductor of the supply network is carried out in such a way that the voltage with the smaller amplitude results at the AC output of this power section, thereby minimizing the leakage currents and enabling the connection of each inverter without the risk of exceeding the defined maximum leakage current. The method according to the invention provides, without measuring the actual leakage current,To determine the leakage capacitance at the DC input of each inverter and, based on the maximum leakage current and the inverter's connection to the power grid, to adjust the modulation of the power sections of each inverter accordingly and to shift the normally equal amplitudes of the voltages at the power section outputs accordingly or to arrange them asymmetrically. While previously the two power sections of each inverter were modulated to produce opposing alternating voltages with the same amplitude, the control is now asymmetrical.with different voltage amplitudes at the AC outputs of the power units. This minimizes or optimizes the leakage currents, making it possible to connect the inverter without the risk of exceeding the maximum leakage current and triggering the residual current device. This allows the inverter to contribute positively to the resulting energy balance even when it would have to remain deactivated according to the state of the art. Since the capacitive leakage current is usually significantly greater than the ohmic leakage current, the leakage capacitance, i.e., the capacitive component of the complex insulation resistance, is primarily determined and considered, rather than the ohmic component, the insulation resistance. The leakage capacitance or system capacitance can be determined, for example, using the method described in EP 3 879 283 A1.

[0011] The respective connection situation of the inverter to the supply grid lines—i.e., which AC output of the inverter is connected to the neutral conductor and which AC connection of the inverter is connected to the phase connection of the supply grid—can be set manually on the inverter or, under certain circumstances, automatically determined. For example, the connection situation can be determined during installation of the inverter and stored or saved in a suitable manner. The method according to the invention then accesses this stored or saved data and thus determines the connection situation.One way to automatically determine the connection situation is, for example, to measure the voltage in a measuring unit (if present) to determine the insulation resistance, and then connect only one power section to the supply network and measure the voltage in the measuring unit a second time. If the AC voltage component is greater in the second measurement than in the first measurement, it can be concluded that the power section connected to the supply network is connected to the phase connection (L) of the supply network, and the other power section is therefore connected to the neutral conductor (N). Subsequently, the counter test can be carried out with the other power section, i.e. only the other power section is connected to the supply network. The AC voltage components of the voltages in both measurements are then compared again.

[0012] The method according to the invention is particularly applicable to inverter arrangements with several inverters, i.e. more powerful systems.

[0013] Typically, the power sections of each inverter are alternately pulse-width modulated at a specified switching frequency. Other modulation methods that control the two power sections accordingly to produce the desired AC output voltage are also conceivable.

[0014] If, after reaching the maximum leakage current, a specified period of time is waited for, and before feeding the AC voltage into the supply grid, the leakage capacitance at the DC input of each inverter to earth is re-determined and a further check is made to determine whether the defined maximum leakage current is exceeded, then, after this specified period of time, a further check can be made to determine whether the respective inverter in the inverter arrangement can be connected without the risk of tripping the residual current device. For example, the specified period of time can be in the range of several minutes, such as 5 minutes. During this time, changes in the leakage capacitance may occur, which may justify a further check.

[0015] The modulation of the power components of each inverter, depending on the measured leakage capacitance, is preferably varied linearly within specified limits. These limits are defined, for example, by extreme values ​​of the leakage capacitance, which could indicate an incorrect measurement or defective hardware in the inverter configuration.

[0016] If a reserve is taken into account in the specified limits, in addition to the leakage currents of the inverter arrangement, any leakage currents of other devices, for example a washing machine connected to the supply network as a consumer, can also be taken into account.

[0017] When specified upper limits for the measured leakage capacitance of each inverter are reached, the respective inverter preferably remains deactivated or is deactivated. The limits for the leakage capacitance are selected and specified based on experience.

[0018] Advantageously, before at least one inverter begins feeding power into the grid, the ohmic insulation resistance at the DC input of each inverter to ground is also measured and taken into account in addition to the leakage capacitance when modulating the power components. As already mentioned above, the influence of the ohmic insulation resistance, in contrast to that of the leakage capacitance, is usually very small or even negligible. However, by also taking into account the ohmic component of the complex insulation resistance, even more efficient operation of the inverter arrangement can be achieved. The ohmic insulation resistance can be determined, for example, using the method described in EP 3 879 277 A1.

[0019] According to a further feature of the invention, the location of a connection between the DC source and ground can be determined, and the determined location can also be taken into account when modulating the inverter's power components. For example, it may be relevant for the modulation whether the ground fault occurs in the area of ​​the positive or negative connection of the inverter's DC input or between individual DC sources, particularly photovoltaic modules, etc. The location at which the ohmic insulation resistance is connected to the DC source can be determined, for example, together with the determination of the ohmic insulation resistance.

[0020] Advantageously, the inverter is deactivated when the intermediate circuit voltage reaches a predetermined upper limit. The method according to the invention may require a higher intermediate circuit voltage, which reduces the overall efficiency of the inverter arrangement. If a certain intermediate circuit voltage, for example, 500 V, is exceeded, it will be advisable to deactivate the inverter for safety reasons and to avoid reduced efficiency.

[0021] Before and / or during feed-in operation of at least one inverter, harmonics of the AC voltage at the AC output can also be determined and taken into account when modulating the inverter's power components. Since harmonics occur in practice and influence the grid situation and contribute to the leakage current, taking them into account can lead to even better utilization of the inverters in the inverter arrangement and thus an improvement in the overall energy balance. Harmonics in the AC output voltage can be determined, for example, using selective filters or Fourier transformations.

[0022] The identified harmonics can be attenuated, for example, by modulating the power section of each inverter connected to the AC output connected to the neutral conductor of the supply grid, and this attenuation can be corrected accordingly by modulating the power section of each inverter connected to the AC output connected to the phase connection of the supply grid. This represents one method for taking harmonics into account.

[0023] According to a further feature of the invention, the DC voltage at the DC input of each inverter can be measured and taken into account when modulating the power components of each inverter. Taking the DC voltage of the DC source, in particular the photovoltaic module, into account can bring further advantages, particularly in inverter arrangements with multiple inverters, since the control of the individual inverters can also be designed depending on the individual input voltages. However, even with only one inverter, taking the DC voltage at the DC input of the inverter into account during modulation can be advantageous and contribute to further increasing efficiency.

[0024] If the amplitude of the alternating voltage is also determined during the feed-in operation of at least one inverter and is taken into account when modulating the power components of each inverter, harmonics, for example, can be taken into account during the operation of the inverter arrangement and the modulation can be optimally adapted to any changed situations.

[0025] In the case of multiple inverters, the power components of each inverter are preferably modulated depending on the measured system capacity of each inverter. Each inverter is controlled or modulated independently of the other inverters in such a way that the proportionally determined maximum leakage current does not occur for each inverter, allowing each inverter to contribute to the energy balance.

[0026] In an inverter arrangement with several inverters, in the simplest case the defined maximum leakage current can be divided by the number of all inverters in the inverter arrangement.

[0027] The number of inverters in the inverter array can preferably be determined automatically via communication lines between the inverters. If, in addition to the number of inverters, the DC voltage at the input of each inverter is also measured and taken into account, a particularly yield-optimized variant of the invention can be implemented.

[0028] The object of the invention is also achieved by an inverter arrangement described above, which is designed to carry out the method described above. The functions of the inverter arrangement can be implemented relatively easily in the control device of the inverter using software. For the achievable advantages, reference is made to the above description of the method.

[0029] The present invention is explained in more detail with reference to the accompanying figures, in which: Fig. 1 shows a block diagram of an inverter arrangement with an inverter; Fig. 2 shows the time profiles of the voltages at the AC outputs of the power units and the resulting AC voltage of the inverter and the resulting leakage current according to the prior art; Fig. 3 shows a flowchart illustrating a first embodiment of a method according to the invention for operating an inverter arrangement; Fig. 4 shows the time profiles of the voltages at the AC outputs of the power units and the resulting AC voltage of the inverter and the resulting leakage current for an ideal supply network according to the method according to the invention; Fig. 5 shows the time profiles of the voltages at the AC outputs of the power units and the resulting AC voltage of the inverter and the resulting leakage current for a supply network with harmonics according to the prior art;6 the time courses of the voltages at the AC outputs of the power units and the resulting AC voltage of the inverter and the resulting leakage current in a noisy supply network with untreated harmonics when applying the method according to the invention according to . Fig. 3 ; Fig. 7 a flowchart illustrating a further embodiment of a method according to the invention for operating an inverter arrangement; Fig. 8 the time profiles of the voltages at the AC outputs of the power units and the resulting AC voltage of the inverter and the resulting leakage current in a noisy supply network with damped harmonics in the method according to the invention according to Fig. 6 ; and Fig. 9 a block diagram of an inverter arrangement with several inverters and a battery.

[0030] Fig. 1 shows a block diagram of an inverter arrangement 1 with a single-phase transformerless inverter 2. The inverter 2 is used to convert a direct voltage U DC , which is provided by a DC source 3, in particular a photovoltaic module 3' or a string of several photovoltaic modules 3', into an alternating voltage U AC with a mains frequency f AC at the output of the inverter 2. The inverter 2 contains a DC / DC converter 5 for transforming the direct voltage U DC at the DC input 4 into an intermediate circuit voltage U ZK across an intermediate circuit capacitor C ZK . The DC / DC converter 5 is preferably formed by a step-up converter (boost converter). The intermediate circuit voltage U ZK is converted into the alternating voltage U AC in a DC / AC converter 6.For this purpose, the DC / AC converter 6 has two power sections 7, 8, usually with corresponding semiconductor switches in a bridge circuit (not shown), which are controlled alternately so that an alternating voltage U AC results at the output of the DC / AC module 6. A corresponding control device 9 is provided to modulate the power sections 7, 8. Pulse width modulation with a predetermined switching frequency usually takes place. At the AC output 10 of the power section 7, the voltage U 10 is applied opposite the lower input of the power section 7 (or the negative pole of the intermediate circuit voltage U ZK ), and at the AC output 11 of the power section 8, the voltage U 11 is applied opposite the lower input of the power section 8 (or the negative pole of the intermediate circuit voltage U ZK ).The two AC outputs 10 and 11 of the power units 7, 8 of the DC / AC converter 6 together form the output of the inverter 2, to which the AC voltage U AC is applied and which is connected to the single-phase supply network 12. In the example shown, the AC output 10 of the power unit 7 is connected to the phase connection L and the AC output 11 of the power unit 8 is connected to the neutral conductor N of the supply network 12. Output chokes 13 can be arranged in the cables at the output of the inverter 2. In addition, an AC isolator 17 is located at the output of the inverter 2, with which the inverter 2 can be disconnected from the supply network 12. An RCD 14 is arranged between the inverter 2 and the supply network 12. This RCD measures a total current and trips if a defined maximum leakage current IA,max to earth PE is exceeded.The leakage currents IA result from insulation faults, which are represented by a complex insulation resistance Z iso in the form of a parallel connection of a leakage capacitance C iso and an insulation resistance R iso between the DC sources 3 and earth PE. The location of the insulation fault, i.e. at which point on the DC source 3 the ohmic insulation resistance R iso is connected to the DC source 3, also influences the DC component of the leakage current IA. The ideal arrangement of the insulation fault would be symmetrical between the two connections of the DC input 4 of the inverter 2, as this results in the smallest DC component of the leakage current IA because, when the inverter 2 is feeding into the grid, the voltage to earth PE is absolutely lowest in the middle. Therefore, determining the location of the connection between the DC source 3 and earth PE may be advantageous under certain circumstances.This determination can, for example, be part of the measurement of the ohmic insulation resistance R iso . If the location is known, the DC component of the leakage current IA can be determined using Ohm's law by dividing the measured voltage of the insulation fault to earth PE by the ohmic insulation resistance R iso . The DC component of the leakage current IA reduces the permissible capacitive component of the leakage current IA , which must be taken into account for the modulation of the power units 7, 8.

[0031] To prevent the residual current device 14 from being triggered, before the inverter 2 is connected to the supply grid 12 via the AC isolator 17, the leakage current IA is measured or the leakage capacitance C iso and, if applicable, the ohmic insulation resistance R iso are determined, and the inverter 2 is only connected if the defined maximum leakage current IA,max is not reached. Otherwise, the inverter 2 cannot be connected to the supply grid 12 and the alternating voltage U AC converted from the direct voltage U DC cannot be fed into the supply grid 12 and thus cannot contribute to the energy balance. This naturally entails yield losses for the operator of the inverter arrangement 1, which are particularly significant for large and powerful systems.

[0032] According to the invention, the modulation of the power units 7, 8 of each inverter 2 is adapted, taking into account the determined leakage capacitance C iso and the defined maximum leakage current IA,max and the determined connection situation, such that the amplitudes of the voltages U 10 , U 11 at the outputs 10, 11 of the power units 7, 8 are different or asymmetrical. As a result, the leakage current IA is minimized or optimized (IA,opt ), and it is therefore possible to connect the inverter 2 without the risk of exceeding the maximum leakage current IA,max and triggering the residual current device 14. Thus, the respective inverter 2 can make a positive contribution to the resulting energy balance even where it would have to remain deactivated according to the state of the art.

[0033] Fig. 2 shows the time curves of the voltages U 10 , U 11 at the AC outputs 10, 11 of the power sections 7, 8 and the resulting AC voltage U AC of the inverter 2 and the resulting leakage current IA in the state of the art. The AC output 10 of the power section 7 of the inverter 2 is designed according to the block diagram of the inverter arrangement 1 according to Fig. 1 connected to the phase connection L and the AC output 11 of the power section 8 of the inverter 2 is connected to the neutral conductor N of the supply network 12. The modulation of the power sections 7, 8 of the DC / AC converter 6 of the inverter 2 occurs in such a way that the two voltages U 10 and U 11 are output in opposite directions with the same amplitude. The maximum possible amplitude of the voltages U 10 and U 11 would be the maximum intermediate circuit voltage U ZK - here 500V. As shown, an amplitude of 325V was selected. The amplitude of the alternating voltage U AC is also 325V - i.e. the voltage of the supply network 12. This results in a leakage current IA (here, for example, with an amplitude of 50 mA and an effective value of 35 mA), which is not affected by the modulation.

[0034] In Fig. 3 is a flow chart illustrating a first embodiment of a method according to the invention for operating an inverter arrangement 1. According to step 101, the leakage capacitance C iso and possibly also the ohmic insulation resistance R iso are determined using a known method and according to step 102, the capacitive component of the insulation resistance X iso is determined from this via X iso = 1 2 ⋅ π ⋅ f AC ⋅ C iso calculated from the leakage capacitance C iso and the mains frequency f AC.

[0035] According to step 103 of the method, the defined maximum leakage current IA,max is determined. In the case of an inverter 2 and a residual current device 14, the maximum leakage current IA,max is determined by this residual current device 14, since it is designed for a specific maximum leakage current IA,max - for example, 30 mA. In the case of multiple inverters 2 within the inverter arrangement 1 (see Fig. 9 ), the defined maximum leakage current IA,max is divided, for example, by the number i of inverters 2 in order to determine the maximum leakage current per inverter 2. But here too, the max. leakage current IA,max , which is determined by the residual current device 14, applies.

[0036] According to step 104, the smaller amplitude of the voltages U 10 or U 11 is calculated. This is done by multiplying the maximum leakage current IA,max by the capacitive component of the insulation resistance X iso . For example, with a leakage capacitance C iso =1 µF and a mains frequency f AC =50 Hz, this results in an X iso =3.18 kOhm and with an intermediate circuit voltage U ZK =500 V, the smaller amplitude of the voltages U 10 , U 11 is approx. 75 V. The larger amplitude of the voltages U 10 , U 11 is obtained by subtracting the smaller amplitude from the AC voltage of 325 V of the supply network 12 (as the product of the square root of 2 and the voltage of 230 V), which is approx. 250 V. However, since a certain tolerance (for example + / - 10%) is realistic in the supply network 12, the voltage of the supply network 12 can also be measured continuously in order to obtain a current value. Accordingly, the amplitude of the AC voltage U AC must also be 325V orcorrespond to the current measured value. The alternating voltage U AC corresponds to the sum of the voltages U 10 and U 11 at the AC outputs 10, 11 of the two power units 7, 8 of inverter 2.

[0037] At this point in time, the connection situation has not yet been determined, i.e., it is not yet known which AC output 10, 11 is connected to the phase conductor L or the neutral conductor N of the supply network 12. However, this is not relevant for the basic determination at this point in time.

[0038] According to query 105, a check is performed to determine whether the determined larger amplitude of the voltages U 10 , U 11 is less than half the intermediate circuit voltage U ZK / 2. If this is the case, feeding is possible. If this is not the case, feeding is not possible. With an intermediate circuit voltage U ZK of 500 V (amplitude), the larger amplitude of the voltages U 10 or U 11 may not exceed 250 V.

[0039] If this condition - i.e. the amplitude of the larger amplitude of the voltage U 10 or U 11 does not exceed half the intermediate circuit voltage U ZK / 2 - is met, the connection situation is first determined according to step 106, i.e. the determination of which AC output 10, 11 of the inverter 2 is connected to which line L, N of the supply network 12.

[0040] The AC isolator 17 is located at the output of inverter 2 and includes at least one relay per conductor. These relays can be controlled accordingly, for example, by the control device 9 and / or by the power units 7, 8. As soon as the relay in the phase conductor L is closed, a sufficiently large AC voltage component at 50 Hz can be measured against earth PE – for example, by the control device 9. This is not the case when the relay in the neutral conductor N is closed. In this way, for example, the connection situation can be determined automatically. Typically, the contacts of the relays of the AC isolator 17 are checked regularly for sticking. The automatic determination of the connection situation can also take place during this test.

[0041] According to step 107, a corresponding modulation of the power units 7, 8 of each inverter 2 then takes place. The power unit 7 on the neutral conductor N is modulated such that the voltage U 10 with the smaller amplitude is produced at output 10. The power unit 8 on the phase conductor L is modulated such that the voltage U 11 with the higher amplitude is produced at output 11. This ensures that the leakage current IA does not exceed the maximum leakage current IA,max and therefore the residual current device 14 is not triggered. This is provided that the leakage current IA is within a range in which a reduction below the maximum leakage current IA,max is possible through the modulation of the power units 7, 8.

[0042] Accordingly, according to step 108, the inverter arrangement 1 is operated in such a way that the power units 7, 8 output different voltages U 10 , U 11 at the AC outputs 10, 11, so that the maximum leakage current IA,max is not exceeded and the inverter 2 is operated with an optimized leakage current IA,opt.

[0043] If the above condition of query 105 is not met, no feed-in to the supply grid 12 takes place according to step 109, by not connecting the respective inverter 2 to the supply grid 12.

[0044] In this case, according to step 110, preferably a predetermined time period Δt, for example, 5 minutes, is waited for and the determination of the leakage capacitance C iso is started again according to step 101. During the time period Δt, an attempt can also be made to increase the intermediate circuit voltage U ZK , for example, with a booster from 500 to 550 V (depending on the intermediate circuit capacitors used and the inverter topology).

[0045] Fig. 4 shows the time profiles of the voltages U 10 , U 11 at the AC outputs 10, 11 of the power units 7, 8 and the resulting AC voltage U AC of the inverter 2 and the resulting leakage current IA in the method according to the invention under the assumption of an ideal supply network 12 without harmonics. According to the invention, before feeding the AC voltage U AC into the supply network 12, the leakage capacitance C iso at the DC input 4 of each inverter 2 against earth PE is determined and calculated according to the flow diagram from Fig. 3 the voltages U 10 , U 11 of the power units 7, 8 are determined and the power units 7, 8 are modulated accordingly, so that the voltages U 10 , U 11 result with the corresponding different amplitudes.

[0046] In the illustrated embodiment according to Fig. 1 The AC output 10 of the power section 7 of the inverter 2 is connected to the phase connection L and the AC output 11 of the power section 8 of the inverter 2 is connected to the neutral conductor N of the supply network 12. Accordingly, the voltage U 11 with the smaller amplitude is output by the power section 8 and the voltage U 10 with the larger amplitude is output by the power section 7. As shown, the amplitude of the voltage U 10 is approximately 250 V and the amplitude of the voltage U 11 is approximately 75 V. For the alternating voltage U AC, this results in an amplitude of 325 V (230 V times the square root of 2), which corresponds to the amplitude of the supply voltage 12.

[0047] As previously described, the lower voltage fluctuations on the neutral conductor N have a positive effect on the level of the leakage current IA .

[0048] The asymmetric modulation of the power units 7, 8 results in an optimized leakage current IA,opt , which is significantly lower than the leakage current IA of the state of the art ( Fig. 2 ) has a lower amplitude. In the example shown, the amplitude of the optimized leakage current IA,opt is approximately 23 mA, or corresponding to an effective value of approximately 16 mA. A conventional residual current device 14, for example, trips at a maximum leakage current IA,max of 30 mA. Accordingly, the optimized leakage current IA,opt can be kept low using the method according to the invention before and during operation of the inverter 2, so that it remains below the maximum leakage current IA,max, as long as the conditions permit (low insulation resistance, high leakage capacitance, little grid interference, hardly any harmonics, etc.). Accordingly, the inverter 2 can feed into the supply grid 12 for longer (switch on earlier, switch off later) and thus contribute more to the energy balance.

[0049] In Fig. 5 The time profiles of the voltages U 10 , U 11 at the AC outputs 10, 11 of the power units 7, 8 and the resulting AC voltage U AC of the inverter 2 and the resulting leakage current IA are shown in the prior art for a supply network 12 with harmonics. The harmonics result from interference in the supply network 12, which is superimposed on the voltages U 10 , U 11. Accordingly, these high-frequency interferences also superimpose the leakage current IA, whereby its amplitude increases - as shown to approximately 80 mA or an effective value of approximately 43 mA, since the voltage fluctuations of the voltage U 11 at the AC output 11 of the power unit 8 are correspondingly amplified.

[0050] Fig. 6 shows the time profiles of the voltages U 10 , U 11 with harmonics and the resulting optimized leakage current IA,opt in a noisy supply network 12 with harmonics when applying the method according to the invention according to Fig. 3 The amplitudes of the voltages shown correspond to those from Fig. 4 , i.e. 250V for U 10 , 75V for U 11 and 325V for U AC . The time courses differ from those of the Fig. 4 by the superposition of high-frequency harmonics on the supply network 12, which are also transferred to the voltages U 10 and U 11. For example, the frequency of the harmonics here is f OW = 350 Hz. The superposition results, among other things, in a larger fluctuating voltage U 11 on the neutral conductor N, which causes a larger fluctuation in the amplitude of the leakage current IA. As shown here, the amplitude is approximately 60 mA or the effective value is approximately 33 mA. In this case, the defined maximum leakage current IA,max , at which the residual current device 14 trips, is exceeded. However, the method according to the invention still shows its effectiveness, because compared to the prior art according to Fig. 5 the amplitude of the leakage current IA would fluctuate even more strongly5.

[0051] In Fig. 7 is a flowchart illustrating a further embodiment of a method according to the invention for operating an inverter arrangement 1. Compared to the flowchart according to Fig. 3 A step 111 has been added here, in which harmonics of the alternating voltage U AC at the AC output 10, 11 are determined before and / or during the feed-in operation of the at least one inverter 2 and are taken into account in the modulation of the power units 7, 8 of the inverter 2. The network analysis according to step 111 can be carried out by determining the harmonics at certain frequencies f OW in step 112. By taking into account such harmonics that frequently occur in reality, the method according to the invention can be further improved - as can be seen from the Fig. 8 which ultimately leads to an even better energy balance.

[0052] Fig. 8 shows the time profiles of the voltages U 10 , U 11 at the AC outputs 10, 11 of the power units 7, 8 and the resulting AC voltage U AC of the inverter 2 and the resulting leakage current IA in a noisy supply network 12 with damped harmonics and the optimized leakage current IA in the inventive method according to claim 9. The determined harmonics are correspondingly damped by modulating the power unit 7 of the inverter 2 that is connected to the neutral conductor N of the supply network 12, as can be seen from the time profile of the voltage U 11. This damping is correspondingly corrected by modulating the power unit 8 of the inverter 2 that is connected to the AC output 10 that is connected to the phase connection L of the supply network 12. The correction orDamping is achieved by transferring the high-frequency components of voltage U 11 to voltage U 10 through appropriate filters in power sections 7 and 8. This is evident from the double amplitude of the superimposed oscillation on voltage U 10 compared to voltage U 10 of . Fig. 7 Accordingly, the voltage U 11 no longer contains any high-frequency components.

[0053] The result is an optimized leakage current IA with a lower overlay of harmonics, which is why the leakage current IA remains below the maximum leakage current IA,max, as long as the ambient conditions (low insulation resistance, high leakage capacitance, little mains interference, hardly any harmonics, etc.) permit this. As shown here, the amplitude is approximately 40 mA, or the effective value is approximately 23 mA. This results in an even longer power supply.

[0054] Fig. 9 shows the block diagram of an inverter arrangement 1 with two inverters 2 and a battery 16 for temporarily storing the energy generated. More than two inverters 2 can also be arranged in parallel in the inverter arrangement 1 and connected jointly to the supply network 12. The control devices 9 can be connected to one another via a communication line 15. The defined maximum leakage current IA,max , at which the common residual current circuit breaker 14 of the inverter arrangement 1 trips, can in the simplest case be divided by the number i of inverters 2, here 2, and for each inverter, half the defined maximum leakage current IA,max / 2 can be used as the basis or new maximum leakage current IA,max ' for this inverter 2 and the inventive method (according to Fig. 3 or Fig. 7 ) for each inverter 2 with this maximum leakage current IA,max ' = IA,max / 2.

[0055] Another possibility is to allocate the maximum leakage currents IA,max ' of the individual inverters 2 according to their rated power. Inverters 2 of inverter arrangement 1 with a higher rated power would then be allocated a larger share of the total maximum leakage current IA,max than inverters 2 with a lower rated power.

[0056] In a yield-optimized variant of an inverter arrangement 1 with multiple inverters 2, it could also be calculated for each inverter 2 how much nominal power it can convert per mA of leakage current IA. This will be different for each inverter 2 due to different leakage capacitances C iso. The maximum total leakage current IA,max is assigned to the inverter 2 with the lowest leakage capacitance C iso per nominal power. If the total quota has not yet been exhausted, the next best inverter 2 is used, and so on. This allows maximum yield to be achieved for a defined maximum total leakage current IA,max.

[0057] Finally, in inverter arrangements 1 with several inverters 2, other aspects of the individual inverters 2, such as service life, operating hours or the like, can also play a role and be taken into account when applying the method according to the invention (according to Fig. 3 or Fig. 7 ) should be taken into account.

[0058] The method according to the invention (according Fig. 3 or Fig. 7 ) for operating an inverter arrangement 1 with at least one, preferably several, inverters 2 minimizes the problem of deactivating the inverters 2 if the defined maximum leakage currents IA,max are exceeded by minimizing the leakage currents IA through special modulation of the power units 7, 8 of the inverters 2, taking into account the leakage capacitance C iso and the connection situation. Since this allows the feed-in time and feed-in power to be increased, this results in a better energy balance and, consequently, higher yields for the operator of the inverter arrangement 1.

Claims

1. A method for operating an inverter assembly (1) having at least one inverter (2) for converting respectively a direct voltage (UDC) into an alternating voltage (UAC) with a mains frequency (fAC), wherein in each inverter (2) the direct voltage (UDC) of a DC source (3) connected to a DC input (4) is transformed via a DC / DC-converter (5) to an intermediate circuit voltage (UZK), and the intermediate circuit voltage (UZK) is converted in a DC / AC-converter (6) having two power units (7, 8), each connected to an AC output (10, 11), which are mutually modulated by a control device (9) of the respective inverter (2), into the alternating voltage (UAC), which alternating voltage (UAC) corresponds to the total of the opposite voltages (U10, U11) at the AC outputs (10, 11) of the power units (7, 8) opposite the negative pole of the intermediate circuit voltage (UZK), and the alternating voltage (UAC) is fed into a supply network (12) connected to the AC outputs (10, 11), characterised in that before feeding the alternating voltage (UAC) into the supply network (12), the leakage capacitance (Ciso) to ground (PE) is determined at the DC input (4) of each inverter (2), and it is determined which AC output (10, 11) is connected to which line (L, N) of the supply network (12), and, taking into consideration the determined leakage capacitance (Ciso) and a defined maximum leakage current (IA,max) and the determined connection situation at the AC output (10, 11), the power units (7, 8) of each inverter (2) are asymmetrically modulated by the control device (9) in such a way that the amplitudes of the voltages (U10, U11) at the AC outputs (10, 11) of the power units (7, 8), which correspond to the alternating voltage (UAC) to be fed in the supply network (12), differ in size, the modulation of the power unit (7) of the DC / AC-converter (6) connected to the phase connection (L) of the supply network (12) taking place in such a way that the voltage (U10) with the greater amplitude comes about at the AC output (10) of this power unit (7), and the modulation of the power unit (8) of the DC / AC-converter (6) connected to the neutral conductor (N) of the supply network (12) taking place in such a way that the voltage (U11) with the lower amplitude comes about at the AC output (11) of this power unit (8), as a result of which the leakage currents (IA) are minimised and it is possible to connect each inverter (2) without the risk of exceeding the defined maximum leakage current (IA,max).

2. The method according to claim 1, characterised in that in the event that the maximum leakage current (IA,max) is reached, a predetermined period of time (Δt) is waited and, before feeding the alternating voltage (UAC) into the supply network (12), the leakage capacitance (Ciso) to ground (PE) at the DC input (4) of each inverter (2) is determined again and it is checked again whether the defined maximum leakage current (IA,max) is exceeded.

3. The method according to claim 1 or 2, characterised in that the modulation of the power units (7, 8) of each inverter (2) is changed linearly within predetermined limits as a function of the measured leakage capacitance (Ciso).

4. The method according to any one of claims 1 to 3, characterised in that upon reaching predetermined upper limit values (Ciso_max) for the measured leakage capacitance (Ciso) of each inverter (2), the respective inverter (2) remains deactivated or is deactivated.

5. The method according to any one of claims 1 to 4, characterised in that, before the feed-in operation of the at least one inverter (2), the ohmic insulation resistance (Riso) to ground (PE) is measured at the DC input (4) of each inverter (2) and is taken into consideration in addition to the leakage capacitance (Ciso) when modulating the power units (7, 8).

6. The method according to claim 5, characterised in that the location of a connection of the DC source (3) to the ground (PE) is determined and the determined location is taken into consideration when modulating the power units (7, 8) of the inverters (2).

7. The method according to any one of claims 1 to 6, characterised in that before and / or during the feed-in operation of the at least one inverter (2), harmonics of the alternating voltage (UAC) at the AC output (10, 11) are determined and taken into consideration when modulating the power units (7, 8) of the inverter (2).

8. The method according to claim 7, characterised in that the determined harmonics are attenuated by modulating that power unit (7) of each inverter (2), which is connected to the AC output (11), which is connected to the neutral conductor (N) of the supply network (12), and this attenuation is correspondingly corrected by modulating that power unit (8) of each inverter (2), which is connected to the AC output (10), which is connected to the phase connection (L) of the supply network (12).

9. The method according to any one of claims 1 to 8, characterised in that during the feed-in operation of the at least one inverter (2), the amplitude of the alternating voltage (UAC) is determined and taken into consideration when modulating the power units (7, 8) of each inverter (2).

10. The method according to any one of claims 1 to 9, characterised in that, in the case of a plurality of inverters (2), the power units (7, 8) of each inverter (2) are modulated as a function of the measured system capacitance (Ciso) of each inverter (2).

11. The method according to claim 10, characterised in that the defined maximum leakage current (IA, max) is divided by the number (i) of all inverters (2) of the inverter assembly (1).

12. The method according to claim 10 or 11, characterised in that the direct voltage (UDC) is measured at the DC input (4) of each inverter (2) and is taken into consideration when modulating the power units (7, 8) of each inverter (2).

13. The method according to any one of claims 10 to 12, characterised in that the number (i) of inverters (2) of the inverter assembly (1) is determined via communication lines (15) between the inverters (2).

14. An inverter assembly (1) having at least one inverter (2) for converting respectively a direct voltage (UDC) into an alternating voltage (UAC) with a mains frequency (fAC), wherein each inverter (2) has a DC / DC-converter (5) for transforming the direct voltage (UDC) of a DC source (3) connected to a DC input (4) into an intermediate circuit voltage (UZK), a DC / AC-converter (6) having two power units (7, 8) for converting the intermediate circuit voltage (UZK) into the alternating voltage (UAC), wherein every inverter (2) comprises a control device (9) for modulating the power units (7, 8) of each inverter (2), characterised in that the control device (9) is configured to carry out the method according to any of one of claims 1 to 13.