Method and circuitry for detecting arcs and photovoltaic (PV) inverter having corresponding circuitry

Abstract

This application describes a method for detecting an arc (9) in a direct current circuit (6) comprising a DC load (3), a DC power source (2) supplying the DC load (3), and a circuit arrangement (1) disposed between the DC power source (2) and the DC load (3). Power flow between the DC power source (2) and an output (8) of the circuit arrangement (1) is inhibited by a switching unit (10) through periodic interruptions such that power flow is enabled during an active time window (31) having a first period Δt1, and power flow is inhibited during an inactive time window (32) having a second period Δt2. Detection of a current I characterizing a power flow P and / or a voltage U characterizing the power flow P in two consecutive active time windows 31 and a comparison of the detected values ​​of the current I and / or voltage U in an active time window 31 with corresponding detected values ​​from the previous active time window 31 can indicate an arc 9 if those values ​​in the active time window 31 differ from the corresponding values ​​in the previous active time window 31 by more than a threshold value. The present application also describes a circuit arrangement 1 for detecting an arc 9 and a photovoltaic (PV) inverter 40 including such circuit arrangement 1.

Description

[Technical field]

[0001] The present invention relates to a method and circuitry for detecting arcs, and to a photovoltaic (PV) inverter including such circuitry. [Background technology]

[0002] Arcs can cause considerable damage to electrical systems, which can result, for example, in fires in the electrical system as well as in the building that contains it. Personal injuries are also possible. For this reason, arc detection is an important requirement in electrical systems. This is especially true when the electrical system has direct current (DC) circuits, since unlike alternating current (AC) circuits, in DC circuits, burning arcs can usually be extinguished, but there are no zero crossings of voltage that can at least facilitate the quenching of the arc.

[0003] A distinction is usually made between two different types of arcs. A series arc occurs when there is a contact fault, for example when a cable or plug is loose, but the contacts are so close together that there is a locally high electric field intensity in the intermediate area between them. An example of a series arc is also a switching arc, which occurs between the contacts of a relay during a switching operation. A parallel arc is much rarer and occurs, for example, when the positive and negative lines of an electrical system are close to each other and their insulation is defective. Here too, there may be a locally high electric field intensity between the damaged conductors. The power losses transformed into a series arc are usually significantly lower than those transformed into a parallel arc. Thus, a series arc between a power source and a load supplied by the power source causes a relatively small change in the power transmitted. In contrast, a parallel arc causes a large change in the power supplied by the power source and ultimately consumed by the load.

[0004] Conventional methods for detecting arcs usually use the noise spectrum emitted by the arc. When there is an arc in an electrical system, a high-frequency interference voltage or current is superimposed on the effective voltage or current. The high-frequency interference voltage and / or current can be electrically filtered and detected on conductors that also carry an effective current or voltage. If a certain interference voltage and / or a certain interference current is exceeded, an arc is signaled. Such a method is described, for example, in WO 2012 / 116722 A1. To increase the sensitivity of the detection, it is also possible to simultaneously detect many different frequencies in the transmitted noise spectrum.

[0005] Another known method for detecting an arc uses the abrupt change in voltage and / or current that accompanies the formation of an arc. In this case, the current and / or voltage are usually monitored with high time resolution for the detection of the abrupt change. If necessary, this method can also be used to ascertain how the observed change in current and / or voltage differs from the normal fluctuations in the power consumption of the electric load. If the observed change in current and / or voltage differs significantly from the normal consumption fluctuations of the grid, an arc is indicated. Such a method is disclosed, for example, in WO 2011 / 017721 A1.

[0006] Finally, WO 2013 / 171329 A1 discloses a method for arc detection that combines the detection of a noise signal with the detection of a sudden change in current, whereby it is possible to provide the temporal occurrence of the detected noise signal as well as the temporal occurrence of the detected current change, using a timestamp in each case. By comparing both timestamps, an additional plausibility check regarding the presence of an arc can be confirmed.

[0007] All methods typically use relatively complex and therefore costly measurement systems, the more complex the measurement system, the more reliably a possible arc is detected or the more reliably a false signaling of an arc and associated deactivation of the electrical system is avoided.

[0008] DE 10225259 B3 discloses an electrical plug connector with main contacts which are connected during the plug withdrawal process and lagging auxiliary contacts which are connected in parallel with them. When the connector is withdrawn, the main contacts are necessarily disconnected first and the auxiliary contacts last from the mating part. To avoid arcs associated with the withdrawal process, a semiconductor switch connected in series with the auxiliary contacts is periodically switched on and off between the disconnection of the main and auxiliary contacts. Summary of the Invention

[0009] The present invention aims to identify a method for detecting arcs, in particular series arcs, in DC circuits that can be implemented relatively inexpensively. In particular, the intention here is to avoid complex detection systems as far as possible, and to use relatively simple and inexpensive measurement techniques instead. It is also an object of the present invention to identify a circuit arrangement suitable for implementing the method, as well as a photovoltaic (PV) inverter comprising such a circuit arrangement.

[0010] The object of disclosing a method for arc detection of the type mentioned at the beginning is achieved according to the invention by the features of independent claim 1. The object of specifying a circuit arrangement suitable for carrying out the method is achieved according to the invention by the features of independent claim 14. The object of disclosing a PV inverter comprising such a circuit arrangement is achieved by the features of independent claim 17. Advantageous modifications of the method are set out in the dependent claims 2 to 13, advantageous embodiments of the circuit arrangement are set out in the dependent claims 15 and 16 and advantageous embodiments of the PV inverter are set out in the dependent claims 18 and 19.

[0011] The method according to the invention aims to detect an arc in a direct current (DC) circuit comprising a DC load, a DC power supply supplying the DC load, and a circuit arrangement arranged between the DC power supply and the DC load, where the circuit arrangement comprises an input with two input terminals for connecting the DC power supply, an output with two output terminals for connecting the DC load, and a switching unit arranged between the input and the output, the switching unit having and being operable in at least two operating modes. In particular, the switching unit optionally detects an output voltage U applied to the output of the circuit arrangement. out In relation to the above, in a first operating mode BM1, a power flow P between the DC power source and the output of the circuitry is enabled, and in a second operating mode BM2, a power flow P between the DC power source and the output of the circuitry is disabled. -First period Δt where P ≠ 0 1 and in the second period Δt 2 periodically interrupting the power flow P between the DC power source and the output of the circuit arrangement by the switching unit such that the power flow P is inhibited in an inactive time window having a time period of - detecting a current I characterizing a power flow P and / or a voltage U characterizing a power flow P during two consecutive active time windows; - comparing detected values ​​of the current I and / or the voltage U of an active time window with corresponding detected values ​​from a previous active time window; - transferring the arc if the values ​​of the current I and / or the voltage U in an active time window differ from the corresponding values ​​in the previous active time window by more than a threshold value.

[0012] The arc to be detected may in particular be a series arc. The circuit arrangement may be a full or partial part of an electrical device such as an energy conversion unit, for example an inverter. The DC load may be embodied as a DC load in the conventional sense and for the present purposes includes only DC consumers. However, the circuit arrangement may comprise a DC-AC converter connected at its output to an AC voltage (AC) load. The AC load may be an electrical device supplied by an AC voltage. However, as an alternative to this, the AC load may be an energy supply grid, where the DC-AC converter supplies AC power converted from DC power. A periodic interruption of a power flow P means a repetitive interruption of the power flow P that satisfies a rule that may occur periodically, but does not have to occur periodically. If the interruption occurs periodically, then the first period Δt 1 and the second period Δt 2 The sum of is constant, i.e., Δt 1 +Δt 2 = constant, whereas in the case of aperiodic interruptions the sum is not constant. If the power fluctuations of the consumption of the DC load can be estimated to be negligibly small, the power flow P can be interrupted periodically. If this cannot be guaranteed, periodic interruptions of the power flow usually occur aperiodic. The current I to be detected, which characterizes the power flow P, may be the current flowing between the DC source and the input of the circuit arrangement. Alternatively, the current I may also be the current flowing between the input of the circuit arrangement and the switching unit or between the switching unit and the output of the circuit arrangement. The detected voltage U may be the voltage applied to the input of the circuit arrangement and / or the voltage applied to the output of the circuit arrangement. It is within the scope of the invention to detect only one current, but alternatively it is also within the scope of the invention to detect several currents which characterize the power flow P between the DC source and the output. Alternatively or additionally, it is possible to detect only one voltage, but also several voltages which characterize the power flow P between the DC source and the output of the circuit arrangement.

[0013] The invention uses the effect that the power flow P, which normally occurs continuously in the DC circuit between the DC source and the DC load, is transformed by the switching unit into a power flow P which occurs discontinuously. This applies at least to a first region of the DC circuit arranged between the DC source and the output of the circuit arrangement. Here, it is possible, but not absolutely necessary, for the power flow P to be interrupted everywhere in the first region of the DC circuit - i.e. over the entire route between the DC source and the output of the switching arrangement. Instead, it is sufficient if a periodic interruption of the power flow occurs only in the subsection between the DC source and the output of the circuit arrangement. In particular, for example, if the power flow is interrupted only in the subsection between the DC source and the output, charging processes of the capacitances present there, which are still decaying, can occur in the remaining subsections of the first region of the DC circuit. However, by inhibiting the power flow in the subsection of the first region of the DC circuit, the power flow in the remaining subsections in the first region of the DC circuit is correspondingly reduced at least significantly. In a second region of the DC circuit arranged between the output of the circuit arrangement and the DC load, the power flow may be either discontinuous or continuous.

[0014] Since there is no power flow from the DC source to the output of the circuit arrangement in the deactivated second time window, zero crossings of the current I are intentionally induced in the DC circuit, but at least in the first region of the DC circuit and at least in a subsection of the first region of the DC circuit that extends from the DC source to the output of the switching arrangement. In summary, the zero crossings are applied to the entire first region of the DC circuit between the DC source and the DC circuit output of the circuit arrangement, and the power flow P is at least significantly reduced in all cases if not completely interrupted in all cases. At the same time, energy is withdrawn from a series arc that may be present in the first region of the DC circuit. Due to the withdrawal of energy, the arc cannot continue to burn, but is instead quenched, at least temporarily, within the deactivated time window. If in an active time window following the deactivated time window a power flow P from the DC source to the output of the circuit arrangement occurs again, the risk that this power flow P will also be accompanied by a burning arc is subsequently significantly reduced compared to the previous active time window, in which the arc was still burning. In other words, if an arc is burning in the earlier of two active time windows, then there is a high probability that an arc will no longer be present in the later of two active time windows, with the probability increasing as the inactive time window or the second period of the inactive time window between the two active time windows increases.

[0015] According to the invention, a respective current I and / or voltage U is detected in each of the two time-offset active time windows, which are assigned to the power flow P of the respective active time window. The electrical variables of the current I and / or voltage U change depending on the presence of an arc in the DC circuit, since the power loss consumed by the arc can also be assigned to the arc. A comparison of the detected values ​​of the current I and / or voltage U of the two time-offset active time windows can therefore be used to determine whether one of the two active time windows has an arc as opposed to the other of the two active time windows. This is particularly true in cases where the consumption of the DC load over time is known for other reasons and / or does not change, at least significantly, over time. Thus, in particular, an arc can be transferred if the values ​​of the current I and / or voltage U of a later active time window over time differ from the corresponding values ​​of the current I and / or voltage of a later active time window by more than a predefined threshold value.

[0016] In the method according to the invention, no special requirements are made from the measurement unit for detecting the electrical variable from the current I and / or the voltage U. The measurement unit does not need to have high accuracy or high time resolution, i.e. it does not need to have a high measurement frequency. This is because it is not important to detect abrupt changes of the electrical variable. Instead, it is sufficient to detect only one value of the electrical variable representing the respective active time window, for example the average value of the respective active time window. For this reason, the measurement unit for detecting the electrical variable can be designed very cost-effectively. In many cases, the measurement unit is already present in the DC circuit, which is why no additional costs arise in this case. According to the invention, the power flow P in the DC circuit between the DC source and the output of the circuit arrangement occurs discontinuously, so that no complex detection unit is required for triggering a temporary interruption by other methods of the power flow occurring continuously, even during normal operation of the DC load, semi-individual energy packets. Instead, the method according to the invention operates as if there was an arc, in particular a series arc, in at least a first region of the DC circuit in each of the active time windows which needs to be quenched by interruption, caused by the switching unit, of the power flow P coming from the DC source in the inactive time window. Information as to whether an arc was actually present in a previous active time window results only from detection of the electrical variables of the current I and / or the voltage U in a later active time window in time and comparison of these with the corresponding detected variables of the current I and / or the voltage U of a previous active time window in time.

[0017] In the inactive time window, the DC load can be supplied from an intermediate energy store, for example a capacitance, already present at the input of the DC load, or from a capacitance that is additionally provided and connected in parallel to the output of the circuit arrangement. Furthermore, the second period Δt 2can be selected to be very small so that the capacitance is only slightly discharged, so that under-supply of the DC load and any potentially related imperfect operation of the DC load can be largely avoided.

[0018] In an advantageous embodiment of the method, the switching unit is adapted to generate an output voltage U out , but at least, the second period of the inactivity time window Δt 2 In response to a transmitted arc, power flow P between the DC power source and the output of the circuitry may be interrupted for a period of time that is long compared to . For example, such an interruption occurring in response to a transmitted arc may last for several minutes or until the arc is manually acknowledged. In this manner, a specialist may be provided with an opportunity to identify and repair damage to the DC circuit caused by the detected arc.

[0019] The active time window and the previous active time window, in which the values ​​of the current I and / or the voltage U are compared with each other, do not necessarily have to follow each other directly. Instead, they can be separated from each other by a number of inactive time windows and at least one further active window. However, in an advantageous embodiment of the method, the active time window and the previous active time window can also be separated from each other by just one inactive time window. In this way, changes in the consumption of the DC load that occur in the meantime are less important when comparing the electrical variables.

[0020] In an advantageous embodiment of the method, the second period of the inactivity time window Δt 2The value of the second period Δt can be selected such that it is not only high enough to quench an arc that may be present in the DC circuit, but instead may be high enough to suppress the re-ignition of a previously quenched arc in an active time window that follows the inactive time window in time. The value of the second period Δt can depend in particular on the thermal time constant associated with the arc and in this case may in particular be greater than the corresponding thermal time constant. The thermal time constant represents the decrease in the electrical conductivity of the plasma arc column in time when the arc is no longer supplied with electrical energy, for example because the power flow in the first area of ​​the DC circuit has been interrupted by a switching unit. In particular, if the time for the new application of the voltage U to the arc column after the arc has been quenched is greater than the thermal time constant, the arc cannot be re-ignited. Furthermore, the value of the necessary minimum interruption time of the current flow can also depend on the power converted in the burning arc. Thus, the second period Δt of the inactive time window can depend in particular on the thermal time constant associated with the arc and in this case may be greater than the corresponding thermal time constant. 2 The value of Δt may also be selected in response to the detected values ​​of the current I and / or the voltage U of the immediately preceding active time window. In particular, the second period Δt increases as the value of the detected current I and / or the detected voltage U from the previous time window increases. 2 The test is performed in the second period Δt 2 It was shown that it is sufficient if the value of 0.1 ms to 10.0 ms, preferably 0.3 ms to 4.0 ms, is included.

[0021] In one variant of this method, the power flow P is interrupted between the input and the output of the circuit arrangement by a switching unit. The circuit arrangement may have an input capacitance connected in parallel to its input. In this case, the input current I in and / or the input voltage U applied to the input across the input terminals in Advantageously, the second period Δt of the inactive time window can further be detected. 2 Here, the detected input current I in and / or the detected input voltage U in Specifically, the second period of the inactivity time window Δt 2is, for example, the input current I detected during the inactive time window. in is a further current threshold I TH2 , below which there is no significant further charging of the input capacitance. This can be achieved, for example, by applying an input voltage U in Furthermore, in particular, a further current threshold I TH2 is selected, in particular, so low that if a series arc exists between the DC power source and the input of the circuit arrangement, the series arc cannot continue to burn. This can be determined or verified, for example, empirically. Thus, if the current exceeds the further current threshold I TH2 below this condition ensures that any series arcs that may exist between the DC power source and the input of the circuitry are safely quenched.

[0022] The input current I flows through the input of the circuit configuration. in can of course be detected by a corresponding current sensor. However, this often requires the reservation of an additional current sensor in addition to the voltage sensor that is present there anyway. In one embodiment of the method in which the circuitry has an input capacitance connected in parallel to its input and the power flow P between the input and the output is interrupted, the current flowing between the DC source and the input of the circuitry is at least partially, but also mostly or completely, the input current I flowing at the input. in In this case, during the inactive time window, the input current I in flows into the input, and in particular, the input current I in Here, the input voltage U applied to the input capacitance in It can also be measured by the change over time of the input voltage U applied across the input capacitance, which is also applied across the input terminals. in The time derivative of now provides a measure of the current flowing between the DC power source and the input of the circuit arrangement. In this way, with just one voltage sensor, the input voltage U applied to the input between the input terminals can be measured. inand the input current I in It is possible to determine both.

[0023] In a further variant of the method, the switching unit of the circuit arrangement is formed by a DC-DC converter, in particular a boost converter. Furthermore, the circuit arrangement has an output capacitance connected in parallel to its output. Advantageously, in the active time window, the voltage U applied to the output capacitance out The switching unit switches the first period Δt 1 For example, the first period of the active time window, Δt 1 When the voltage U applied to the output capacitance reaches or exceeds at least 50%, preferably at least 75%, particularly preferably at least 90% of out By doing so, the voltage applied to the output capacitance at the beginning of the active time window, U out In any case, in a DC-DC converter equipped with a freewheeling diode and designed as a boost converter, the voltage applied to the output capacitance U out must be at least as large as the open circuit voltage of the DC power supply. This is because the voltage U applied to the output out This is the only way to ensure that the switching units associated with the circuit arrangement suppress the power flow between the input and the output of the circuit arrangement. Furthermore, in this way it is possible to increase the energy that is intermediately stored in the output capacitance and from which the DC load is also supplied, especially in the immediately adjacent inactive time window. 2 The upper limit of can be increased and the risk of incorrect supply of the DC load, especially during the inactive time window, can be avoided or at least reduced. In contrast, losses of the DC-DC converter occurring due to intermediate storage of energy only occur at the end of the active time window, i.e. during the first period Δt 1t occurs only during a portion of the first time window and not throughout the entire first time window. Additionally, the components providing the output capacitance are loaded with a high voltage value for a small portion of the first period.

[0024] In one embodiment of the method, the arc energy generated by the power loss of the arc expected in the active time window is limited to a predetermined maximum allowable energy value E max The first period of the active time window, Δt 1 In this case, it is assumed that there is an imaginary arc in the current active time window, which does not necessarily exist in reality. However, based on the assumption of an imaginary arc burning in the active time window, the current I in the active time window and the assumed voltage drop Δu of the arc can be selected. LB is used to determine the first period of the active time window, Δt 1 can be estimated by the following formula: TIFF0007661343000001.tif14170

[0025] Depending on the time resolution of the current measurement unit, several detected values ​​of the current I can also be used. In this case, in the denominator of the above formula for the current I, the average current I average can be used. Arc voltage drop ΔU LB There are experimentally determined or empirical values ​​for V. For example, in conventional PV systems, a value of 30V to 40V can be used as a good approximation of the voltage drop across the arc.

[0026] The maximum possible energy value E for detecting an arc in the DC circuit of a photovoltaic (PV) system max is also listed in the relevant standards. max= 200 J. In particular, if the amount of energy delivered in an active time window does not exceed the value 200 J, the PV system can be started again in a simplified manner and without the need for special electrical expertise if an arc was indeed detected in the previous active time window.

[0027] In one embodiment of the method, multiple DC power sources can be connected in parallel to each other via a circuit arrangement in each case to the DC load in order to jointly supply the DC load with their respective power flows P. In doing so, the method for detecting arcs, in particular series arcs, can be executed simultaneously for each of the DC power sources, the inactive time windows assigned to the respective periodic interruptions of the power flows P being preferably arranged to be offset in time from each other when the method is executed for different DC power sources and, if possible, non-overlapping. Advantageously, the periodic interruptions of the power flows P are executed such that at least the overlap in time between the inactive time windows of the various DC power sources is minimized. In this way, a stable operation of the DC load is ensured by minimizing the fluctuations in the total power flow resulting from the sum of the individual power flows P from the multiple DC power sources.

[0028] A circuit arrangement according to the present invention for detecting arcs in a DC circuit comprises: an input having two input terminals for connecting a DC power source and an output having two output terminals for connecting a DC load; a switching unit arranged between the input and the output, the switching unit optionally controlling an output voltage U out a switching unit designed and configured in a first operating mode BM1 to enable a power flow P between the input and the output and in a second operating mode BM2 to inhibit a power flow P between the input and the output, a measuring unit for determining the voltage U applied to the input or output and / or the current I flowing through a connection line from one of the input terminals to one of the output terminals; a control unit for controlling the switching unit and optionally the measuring unit.

[0029] The circuit arrangement is characterized in that the control unit is configured and set in conjunction with the switching unit and the measuring unit to carry out the method according to the invention.

[0030] In order for the switching unit to be able to inhibit the power flow P between the input and the output, the switching unit can have one or more controllable switches, in particular semiconductor switches. Alternatively or additionally, it is also possible for the switching unit to have one or more non-controllable semiconductor switches, for example diodes, in order to inhibit the power flow P. The control unit can also be designed as a separate control unit. However, as an alternative, the control unit can also be designed as a component of a control unit already present in the DC circuit, which results in the advantages already explained in connection with this method.

[0031] In an advantageous embodiment of the circuit configuration, the switching unit may comprise a semiconductor switch arranged in one of the connection lines from one of the input terminals to one of the output terminals. The semiconductor switch may not have an inherently formed or separately formed freewheel diode connected in parallel to the current-carrying terminals of the semiconductor switch. However, as an alternative to this, the semiconductor switch may have a freewheel diode connected in parallel to the current-carrying terminals of the semiconductor switch. In this case, the freewheel diode can be designed either as an inherent freewheel diode or as a separately formed freewheel diode. If the semiconductor switch comprises a freewheel diode, the semiconductor switch may in particular be arranged to be reverse-biased with respect to the normal current direction of the DC circuit.

[0032] The circuit arrangement may comprise an intermediate energy storage connected in parallel to the output, in particular a capacitance. Using the capacitance, the circuit arrangement can intermediately store a limited amount of energy from the DC source in an active time window, with the energy being supplied to the DC load in a later inactive window. In this way, it is possible to supply a DC load during an inactive time window, even if the DC load does not include the possibility of intermediate storage. In this case, the power flow P is discontinuous only in a first area of ​​the DC circuit, while it is continuous in a second area of ​​the DC circuit. According to an embodiment, the circuit arrangement or the switching unit of the circuit arrangement according to the invention can be designed as a DC-DC converter, in particular a boost converter. In this case, the power flow P from the DC source to the output of the circuit arrangement can be interrupted in that the circuit arrangement, in particular the switching unit of the circuit arrangement, is deactivated and the voltage applied to the output of the circuit arrangement is greater than the voltage applied to the input of the circuit arrangement. Experiments have shown that deactivation of several, advantageously up to eight, switching cycles of the DC-DC converter is sufficient.

[0033] A photovoltaic (PV) inverter according to the invention comprises at least one input with two input terminals for connecting a PV string as a DC power source and an output for connecting an AC voltage (AC) grid, a DC-AC converter for converting the DC voltage into an AC voltage, and a circuit arrangement according to the invention, where the input of the circuit arrangement is connected to at least one input of the PV inverter and the output of the circuit arrangement is connected to an input of the DC-AC converter. The inverter may be a single-phase PV inverter with at most one phase terminal and one neutral terminal on the output side. Alternatively, the inverter may be a multi-phase PV inverter with several phase terminals on the output side, in particular three phase terminals on the output side and one neutral terminal on the output side. The control unit of the circuit arrangement can be integrated in the PV inverter or in a higher-level control unit of a PV system comprising the PV inverter. The advantages already mentioned in connection with the method and the circuit arrangement are obtained.

[0034] The method according to the invention for detecting arcs, in particular series arcs, when applied in a PV system in which the DC power supply is constituted by a PV string and the DC load comprises a single-phase DC-AC converter connected at its output side to an AC voltage (AC) grid, comprises detecting a second period Δt 2 The AC current I flows through the output of the DC-AC converter during AC It is advantageous for the periodic interruption of the power flow P to occur such that there are zero crossings of (t) at the time of the power supply. In the case of a single-phase DC-AC converter, power is drawn from an intermediate storage part of the PV inverter, for example from the link circuit capacitance of the DC link circuit, i.e. in a sinusoidal pulse manner, at a frequency corresponding to twice the frequency of the AC grid connected to the PV inverter, and the AC current I generated by the DC-AC converter is AC Proximal to the zero crossing of (t), the lowest power or energy is drawn from the intermediate storage. The inactive time window is the AC current I AC If it overlaps temporarily with the zero crossing of (t), the voltage ripple in the DC link circuit can be minimized.

[0035] In one embodiment of the invention, the PV inverter may have only one input to which the PV string is connected as a DC power source. However, in a further embodiment of the invention, the PV inverter may also have several inputs, each designed to connect a respective PV string as a DC power source. In this case, the inputs may each be connected in parallel to one another via a DC-DC converter to a common DC link circuit. The DC link circuit, which may include a link circuit capacitance, may in turn be connected to the input of the DC-AC converter. In this case, each of the inputs of the PV inverter may be assigned a circuit arrangement according to the invention, each of the inputs of the PV inverter being connected to an input of the circuit arrangement assigned thereto, and the output of each of the circuit arrangements being connected to an input of the DC-AC converter respectively. It is also within the scope of the invention if each of the circuit arrangements is at least partially, possibly completely, formed by one of the DC-DC converters respectively.

[0036] In one embodiment of the PV inverter, the control unit of the circuit arrangement may be included in a higher-level control unit of the PV inverter, where the higher-level control unit is simultaneously configured to control the DC-AC converter and the DC-DC converter of the PV inverter such that the voltage applied to the DC link circuit exceeds the maximum open circuit voltage of the individual PV string during operation of the PV inverter.

[0037] Advantageous embodiments of the invention are specified in the following description and in the dependent claims, the features of which may be applied both individually and in any combination with one another. [Brief description of the drawings]

[0038] In the following, the invention will be explained with the aid of drawings.

[0039] [Figure 1] 1 shows a DC circuit comprising a DC power source, a DC load and a circuit configuration according to a first embodiment of the present invention. [Diagram 2] 4 shows a circuit configuration according to a second embodiment of the present invention. [Diagram 3] 3 shows the operating mode, current and voltage time profiles of the circuit arrangement from FIG. 2 during the method according to the invention. [Figure 4] 1 shows a PV system including a PV inverter according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0040] FIG. 1 shows a DC circuit 6 comprising a direct current (DC) source 2, a DC load 3 and a circuit arrangement 1 according to a first embodiment of the invention. The DC source 2 may have an internal resistance (not explicitly shown in FIG. 1). The circuit arrangement 1 has an input 7 with a first input terminal 7.1 and a second input terminal 7.2, to which the DC source 2 is connected. The circuit arrangement 1 also has an output 8 with a first output terminal 8.1 and a second output terminal, to which the output is connected to the DC load 3. Each of the input terminals 7.1, 7.2 is connected to a corresponding one of the output terminals 8.1, 8.2 via respective connections 23, 24. A switching unit 10 is arranged between the input 7 and the output 8. The switching unit 10 is operable to selectively switch a voltage U applied to the output 8 of the circuit arrangement 1. out In connection with the switching unit 10, the switching unit 10 is designed to enable a power flow P between the DC source 2 and the output 8 during a first operating mode BM1 and to inhibit the power flow P during a second operating mode BM2. For example, a power flow P between the input 7 and the output 8 is allowed by the switching unit 10 during the first operating mode BM1 and prevented during the second operating mode BM2. For this purpose, the switching unit 10 comprises at least one controllable switch, possibly several controllable switches. The switches may be electromechanical switches, but are advantageously semiconductor switches. Furthermore, the switching unit 10 may also comprise one or more non-controllable semiconductor switches, in particular diodes. The switching unit 10 is controlled by the control unit 5. The circuit arrangement 1 also controls an output voltage U applied to the output 8 of the circuit arrangement 1. out and / or the output current I flowing through output 8 out Alternatively or additionally, the measuring unit 4 may be provided with a measuring unit 4 designed to measure an input voltage U applied to the input 7. in and / or the input current I flowing through input 7 in(not shown in FIG. 1 ). The measurement unit 4 is connected to a control unit 5 for communication and control purposes. The circuit arrangement 1 may optionally comprise an output capacitance 11 - shown in dashed lines in FIG. 1 - connected in parallel to its output 8. Alternatively or additionally, the circuit arrangement 1 may comprise an input capacitance connected in parallel to its output 7 (not shown in FIG. 1 ).

[0041] During operation of the circuit arrangement 1, the power flow P between the DC power source 2 and the output 8 of the circuit arrangement 1 is periodically interrupted here by a switching unit 10 - and the output voltage U out Thus, the power flow P between the DC source 2 and the output 8 is interrupted periodically in a first period Δt 1 and a second period Δt 2 Thus, during the active time window 31, a power flow P different from zero occurs between the DC source 2 and the output 8, whereas in the inactive time window 32 the power flow P has the value zero.

[0042] During two successive active time windows 31, the current I and / or the voltage U are here respectively determined by the measurement unit 4 and transmitted to the control unit 5. The current I is calculated based on the input current I in or output current I out The voltage may be the input voltage U in or output voltage U out By the control unit 5 the values ​​of the current I and / or the voltage U of successive active time windows 31 are compared with each other. If the value of the current I and / or the voltage U of a chronologically later active time window 31 deviates from the corresponding value of the current I and / or the voltage U of a previous active time window 31 by more than a threshold value, the control unit 5 is set to transfer an arc 9, in particular a series arc.

[0043] Fig. 2 shows a circuit configuration 1 according to a second embodiment of the present invention. The circuit configuration 1 is similar in some components to the first embodiment already described in Fig. 1, which is why reference is made to the description in Fig. 1 for corresponding features. Mainly, the differences between the second embodiment and the first embodiment are described below.

[0044] According to the second embodiment, the circuit arrangement 1 is a DC-DC converter 21, in particular a boost converter, which converts an input voltage U in The output voltage U applied to output 8 out For this purpose, the switching unit 10 of the circuit arrangement 1 comprises an inductance 22 and a first semiconductor switch S1 connected in series with the inductance 22 in a first connecting line 23 between the first input terminal 7.1 and the first output terminal 8.1. 1 The second semiconductor switch S 2 is connected to the connection point 25 of the inductance 22 through its first terminal, and the first semiconductor switch S 1 The first semiconductor switch S is connected to the second connection line 24 via its second terminal. 1 and a second semiconductor switch S 2 Each of the diodes D1, D2, D3, D4, D5, D6, D7, D8, D9, D10, D11, D12, D13, D14, D15, D16, D17, D18, D19, D20, D21, D22, D23, D24, D25, D36, D37, D48, D59, D61, D62, D70, D71 1 , D 2 It may comprise:

[0045] During the first operating mode BM1, in other words, the first period Δt 1 In the active time window 31 shown in FIG. 3, the circuit configuration 1 operates as a boost converter and the first semiconductor switch S 1 and optionally a second semiconductor switch S 2 By proper clocking of the input voltage U in , the input voltage U in Larger output voltage U out In this case, the output voltage U out is the open circuit voltage U of the DC power source 2 (not shown in FIG. 2) connected to the input 7 0The output voltage U out The energy associated with is temporarily stored in the output capacitance 11 arranged in parallel with the output 8. In the second operating mode BM2, i.e., during the second period Δt shown in FIG. 2 During the inactive time window 32 having 1 , S 2 The clocking of the semiconductor switch S is interrupted by the control unit 5. 1 , S 2 is permanently open during the second operating mode BM2. 2 can last for several clock cycles of the DC-DC converter 21. Here, the voltage U applied to the output 8 out is the open circuit voltage U of the DC power source 2 connected to the input 0 , so that the first semiconductor switch S 1 Related Freewheeling Diode D 1 interrupts the current flow from the first input terminal 7.1 to the first output terminal 8.1 and therefore interrupts the power flow P from the DC power source 2 to the output 8 of the circuit arrangement 1.

[0046] The control unit 5 controls the circuit arrangement 1 to control a first period Δt 1 and the second period Δt for the second operating mode BM1 2 and , which generates a discontinuous power flow P from the DC source 2 to the output 8 of the circuit arrangement 1. The detection and evaluation of the values ​​of the current I and / or the voltage U in two chronologically successive active time windows 31 and the transfer of the arc on the basis of the comparison are performed by the control unit 5, as already explained in connection with FIG.

[0047] 3 shows the time profiles of the current I(t) (middle graph) and the voltage U(t) (lower graph) of the circuit arrangement 1 of FIG. 2 during the operating modes BM1, BM2 (top graph) of the switching unit 10 of the circuit arrangement 1 and during the method according to the invention. The time profiles of the current and the voltage are respectively a function of the variable I(t) detected by the measuring unit 4 at the input 7 of the circuit arrangement 1, for example. in (t), U in (t). In this case, the time profile includes time t 0 4 reflects the behavior of circuit configuration 1 in close temporal proximity to a series arc 9 igniting at

[0048] As shown in the upper graph, the period Δt 1 The active time window 31 of the second operating mode BM2 has a period Δt 2 In the first active time window 31, the current I in (t) is the value I 1 and voltage U in (t) is the value U 1 Take the current I 1 and voltage U 1 The value of Δt results from the power consumption of the DC load 3 connected to the output 8 of the circuitry 1. In the immediately adjacent inactive time window 32, the power flow P between the DC source 2 and the output 8 of the circuitry 1 is 2 The current I flowing through input 7 is interrupted during in (t) takes on a value of 0A, while the voltage U applied to input 7 in (t) is the open circuit voltage U of DC power supply 2 0 Corresponding to the open circuit voltage U 0 is usually the voltage U of the DC source 2 under load due to the always-present internal resistance of the real DC source 2, e.g., a PV generator. 1 In the next active time window 31, in this case the second active time window 31, the current I in (t) is the time t when the arc 9 in the DC circuit 6 first ignites, specifically between the DC power source 2 and the input of the circuit configuration 1.0 corresponds to the value of the previous active time window 31 up to time t 0 From the original value I 1 , U 1 From value I 2 , U 2 The current to t and possibly the voltage also suddenly drops slightly. The sudden drop in current causes a power loss and 0 For example, the series arc 9 burns across a microgap in the faulted conductor. In the adjacent inactive time window 32, the series arc 9 is quenched by again interrupting the power flow P between the DC power source 2 and the output 8 of the circuit arrangement 1. The current I flowing through the input 7 in (t) again takes the value 0 A. Since the microgap is present upstream of the input, the voltage U applied to the input 7 in the second inactive time window 32 in (t) is no longer defined via the DC power supply 2, but rather the voltage U applied to the output 8. out (t) and open semiconductor switch S 1 , S 2 These are defined via the impedance ratio of the voltage U applied to the output 8, out (t) functions as a voltage divider circuit for voltage U in For the profile (t), the voltage U applied to output 8 out It is also necessary to take into account that (t) may be discharged more or less depending on the current that may flow through the DC load 3. In any case, the voltage U applied to the input in (t) is the open circuit voltage U of DC power source 2 0 FIG. 3 shows, by way of example, a voltage U of approximately 0 V in the second inactive time window 32. in In the subsequent third active time window 31, a power flow P through the switching unit 10 is in principle possible, but the power flow P is blocked by the microgap, because the voltage U in (t) and current I in This is because the value of (t) is shown to be negligibly low.

[0049] The control unit 5 here measures the current I in (t) and / or voltage U in (t) with the values ​​of the respective previous active time window 31. When comparing the values ​​between the second active time window 31 and the third active time window 31 in Fig. 3, the control unit 5 determines a difference between these values ​​that exceeds a threshold value and in response transmits the arc 9. In response to the transmission of the arc 9, the switching unit 10 remains in the second operating mode BM2 with the power flow P between the DC source 2 and the output 8 inhibited until manual confirmation by a qualified person.

[0050] Fig. 4 shows a photovoltaic (PV) system 47 with a PV inverter 40 according to an embodiment of the present invention. The PV inverter 40 is here designed as known as a multi-string inverter with several (in this case as an example: two) DC-side inputs 43 with input terminals 43.1 and 43.2 for connecting each PV string 45 as a DC source 2. Each of the inputs 43 is connected in parallel to a link circuit capacitance of a common DC link circuit 42 via a respective DC-DC converter 21, in this case a boost converter. The link circuit capacitance is connected to the input of a DC-AC converter 41. The output of the DC-AC converter 41 is connected to an AC voltage (AC) grid 46 via an output 44 of the PV inverter 40 via a corresponding AC isolation element (not shown in Fig. 4).

[0051] The PV inverter 40 comprises a circuit arrangement 1 for each of the DC-side inputs 43. The circuit arrangement 1, in particular its switching unit 10, is in this case each designed as a DC-DC converter and is formed at least partially by the DC-DC converters 21 of the PV inverter 40 present in any case. The inputs 7 of the circuit arrangement 1 here correspond to the inputs 43 of the PV inverter 40. The outputs 8 of the circuit arrangement 1 here correspond respectively to the outputs of the DC-DC converters 21. The circuit arrangement 1 comprises a common control unit 5 which is part of the central control unit of the PV inverter 40 and which is designed and arranged to control the DC-DC converters 21 and the DC-AC converters 41. The measuring units 4 of the circuit arrangement are also part of the DC-DC converters 21 and are not explicitly shown in FIG. 4.

[0052] The method according to the invention is executed via the control unit 5 simultaneously for each DC source 2, designated as a PV string 45. Each of the PV strings 45 thus has a discontinuous power flow P in its respective DC line between the PV modules and the corresponding DC side input 43 of the PV inverter 40 during normal operation of the PV inverter 40. In this case, during the first period Δt 1 The active time window 31 having a second period Δt 25, alternate with non-active time windows 32 having a non-continuous power flow P in the two PV strings 45. Advantageously, the discontinuous power flow P in the two PV strings 45 occurs here in such a way that there is as little overlap as possible between the non-active time windows 32 of one PV string 45 and the non-active time windows of the other PV string 45, i.e. the non-active time windows are offset as far away from each other as possible. If an arc 9, in particular a series arc, is determined by the control unit 5 in one of the PV strings 45, only that switching unit 10, i.e. the DC-DC converter 21 assigned to that PV string 45 in which the series arc 9 was also detected, is set in the second operating mode BM2 for a longer period and possibly until manual confirmation by a qualified person. In contrast, the respective other PV string 45 can continue to be operated with a discontinuous power flow P in the direction of the associated DC-side input 43. [Explanation of symbols]

[0053] 1 Circuit configuration 2 DC power supply 3 DC load 4 Measurement Units 5. Control Unit 6 Direct current (DC) circuit 7 Input 7.1, 7.2 Input terminals 8 Output 8.1, 8.2 output terminals 9 Arc (Parallel) 10 Switching Unit 11 Output capacity 21 DC-DC converter 22 Inductance 23, 24 Connection wire 25 Connection points 31, 32 hour window 40 Photovoltaic (PV) inverter 41 DC-AC converter 42 DC link circuit 43 Input 43.1, 43.2 input terminals 44 Output 45 PV strings 46 AC Voltage (AC) Grid 47 Photovoltaic (PV) Systems Δt 1 , Δt 2 Time window S 1 , S 2 Solid State Switch BM1, BM2 operation mode D 1 , D 2 diode P power flow U in Input voltage U out Output Voltage I in Input Current I out Output Current

Claims

1. 1. A method for detecting an arc (9) in a direct current (DC) circuit (6) comprising a DC load (3), a DC power source (2) supplying power to the DC load (3), and a circuit arrangement (1) disposed between the DC power source (2) and the DC load (3), comprising: said circuit arrangement (1) comprises an input (7) with two input terminals (7.1, 7.2) for connecting said DC power source (2), an output (8) with two output terminals (8.1, 8.2) for connecting said DC load (3), and a switching unit (10) arranged between said input (7) and said output (8), The switching unit (10) optionally controls the output voltage U ｏｕｔ in a first operating mode BM1 designed to enable a power flow P between the DC power source (2) and the output (8) and in a second operating mode BM2 designed to inhibit the power flow between the DC power source (2) and the output (8), The method comprising: The power flow P occurs during a first period Δt １ and the power flow P is enabled during an active time window (31) having a second period Δt ２ periodically interrupting the power flow P between the DC source (2) and the output (8), in particular between the input (7) and the output (8), by the switching unit (10), such that the power flow P is inhibited during an inactive time window (32) having a period of time that is equal to or greater than the period of time that the switching unit (10) is inactive; - detecting a current I characterizing said power flow P and / or a voltage U characterizing said power flow P during two consecutive active time windows (31); - comparing the detected values ​​of the current I and / or voltage U of the active time window (31) with the corresponding detected values ​​from the previous active time window (31); - reporting the detection of an arc (9) if the values ​​of current I and / or voltage U of the active time window (31) differ from the corresponding values ​​of the previous active time window (31) by more than a threshold value.

2. 2. The method of claim 1, wherein in response to a transferred arc (9), the power flow P between the DC power source (2) and the output (8) is interrupted for a period of at least several minutes, optionally until manual confirmation by the switching unit (10).

3. 3. The method according to claim 1 or 2, wherein the active time window (31) and the previous active time window (31) are separated from each other by exactly one inactive time window (32).

4. The second period Δt ２ is sufficient to quench any arc (9) that may be present in the DC circuit (6) and to inhibit, during the active time window (31) that chronologically immediately follows the inactive time window (32), re-ignition of the arc (9) that was quenched during the inactive time window (32).

5. Said second period Δt ２ The method of claim 4, wherein comprises a value between 0.1 ms and 10.0 ms, preferably between 0.3 ms and 4.0 ms.

6. Said second period Δt ２ The value of is selected depending on the detected values ​​of the current I and / or the voltage U of the previous active time window (31), in particular the second period Δt ２ The method according to any one of claims 1 to 5, wherein increases as the value of the current I and / or the voltage U increases.

7. The circuit arrangement (1) has an input capacitance connected in parallel to its input (7), and further comprises an input current I flowing to the input (7) between the input terminals (7.1, 7.2) during the inactive time window (32). ｉｎ and / or the input voltage U applied to the input (7). ｉｎ is detected, and the second period Δt ２ The detected input current I ｉｎ and / or the detected input voltage U ｉｎ The method according to any one of claims 1 to 5, wherein the method is determined according to

8. The second period Δt ２ is the detected input current I ｉｎ is a further current threshold I ＴＨ２ The method of claim 7, wherein the sigma-based threshold is limited if the sigma-based threshold is less than 0.

9. The circuit arrangement (1) has an input capacitance connected in parallel to its input (7), and during the inactive time window (32), an input current I ｉｎ is the input voltage U applied to the input capacitance ｉｎ The method according to any one of claims 1 to 8, wherein the change in the concentration of the active ingredient is measured over time.

10. The switching unit (10) is formed by a DC-DC converter, in particular a boost converter, and the circuit arrangement (1) has an output capacitance (11) connected in parallel to the output (8), and a voltage U ｏｕｔ The first period Δt １ The method according to any one of claims 1 to 9, wherein the concentration of the ion exchange resin is increased towards the end of

11. The first period Δt １ The value of is determined so that the arc energy generated by the assumed arc power loss during the active time window (31) exceeds a predetermined maximum allowable energy value E ｍａｘ , especially E ｍａｘ 11. The method according to claim 1, wherein the value of .DELTA..times ...

12. The DC power source (2) comprises a PV string (43) and the DC load (3) comprises a single-phase DC-AC converter (41) connected at its output to an AC voltage (AC) grid (45), and the periodic interruption of the power flow P occurs during the second period Δt ２ The AC current I flows through the output of the DC-AC converter (41) during ＡＣ A method according to any one of claims 1 to 11, wherein the zero crossings of (t) are generated such that there are zero crossings of (t).

13. 13. The method according to any one of claims 1 to 12, wherein a plurality of DC power sources (2) are each connected in parallel to the DC load (3) via a switching unit (10) for jointly supplying their respective power flows P to the DC load (3), and the method is performed simultaneously for each of the DC power sources (2) such that temporal overlap of the inactive time windows (32) of the plurality of DC power sources (2) is minimized.

14. A circuit arrangement (1) for detecting an arc (9) in a DC circuit (6), comprising: an input (7) with two input terminals (7.1, 7.2) for connecting a DC source (2) and an output (8) with two output terminals (8.1, 8.2) for connecting a DC load (3); a switching unit (10) arranged between said input (7) and said output (8), said switching unit (10) optionally controlling an output voltage U applied to said output of said circuit arrangement (1); ｏｕｔ a switching unit (10) designed to enable a power flow P between said input (7) and said output (8) in a first operating mode (BM1) and to inhibit a power flow P between said input (7) and said output (8) in a second operating mode (BM2); a measuring unit (4) for additionally determining the voltage U applied to said input (7) or said output (8) and / or the current I flowing in a connection line (23, 24) from one of said input terminals (7.1, 7.2) to one of said output terminals (8.1, 8.2), a control unit (5) for controlling said switching unit (10) and optionally said measuring unit (4), A circuit arrangement (1), characterized in that the control unit (5) in conjunction with the switching unit (10) and the measuring unit (4) is configured and arranged to perform the method according to any one of claims 1 to 13.

15. 15. The circuit arrangement (1) according to claim 14, characterized in that the switching unit (10) comprises a semiconductor switch arranged in one of the connecting lines (23, 24), the semiconductor switch not including a freewheeling diode or the semiconductor switch having an inherent freewheeling diode which is reverse biased with respect to the normal current direction of the DC circuit (6).

16. 16. The circuit arrangement (1) according to claim 14 or 15, characterized in that the circuit arrangement (1) or the switching unit (10) of the circuit arrangement (1) is designed as a DC-DC converter (21), in particular as a boost converter.

17. A photovoltaic (PV) inverter (40), comprising: at least one input (43) with two input terminals (43.1, 43.2) for connecting a PV string (45) as a DC power source (2) and an output (44) for connecting to an AC voltage (AC) grid (46); - a DC-AC converter (41) for converting a DC voltage into an AC voltage; - A photovoltaic (PV) inverter (40) comprising: a circuit arrangement (1) according to any one of claims 14 to 16, the input (7) of which is connected to the at least one input (43) of the PV inverter (40) and the output (8) of which is connected to an input of the DC-AC converter (41).

18. 18. A PV inverter (40) according to claim 17, comprising a plurality of inputs (43) for connecting respective PV strings (45) as a DC power source (2), the inputs (43) being each connected in parallel with one another via a respective DC-DC converter (21) to a common DC link circuit (42), the DC link circuit (42) being connected to an input of the DC-AC converter (41), and comprising a plurality of circuit arrangements (1) according to any one of claims 14 to 16, each of the inputs (43) of the PV inverter (40) being connected to the respective input (7) of the circuit arrangement (1) assigned thereto, and the output (8) of each of the circuit arrangements (1) being connected to the respective input of the DC-AC converter (42).

19. 19. The PV inverter (40) of claim 18, wherein the control unit (5) is configured to simultaneously control the DC-AC converter (41) and the DC-DC converter (21) of the PV inverter (40) such that, during operation of the PV inverter (40), a voltage applied to the DC link circuit (42) is maintained above a maximum open circuit voltage of each of the PV strings (45).

Images