EP2936178A1

EP2936178A1 - Method and device for monitoring a photovoltaic system

Info

Publication number: EP2936178A1
Application number: EP13786161.3A
Authority: EP
Inventors: Christian Strobl; Jochen SEEFRIED; Albert Gerngross
Original assignee: Ellenberger and Poensgen GmbH
Current assignee: Ellenberger and Poensgen GmbH
Priority date: 2012-12-18
Filing date: 2013-10-11
Publication date: 2015-10-28
Also published as: DE102012024728A1; CN104904114B; US20150288331A1; WO2014094929A1; CN104904114A

Abstract

The invention relates to a device (16) and a method (30) for monitoring a photovoltaic system (2) having a number of parallel-connected strings (8) routed towards a common connecting lead (4, 6) for a reverse current (l_R). In the method (30) it is provided for a current flow (ls) to be detected in a preferred direction (12) in each of the strings (8) and for a first total current flow (50) to be created therefrom, for a current flow (IA) to be detected in the preferred direction (12) in the connecting lead (4, 6) and for a second total current flow (54) to be created therefrom, and for the first total current flow (50) to be compared with the second total current flow (54), wherein a reverse current (IR) is recognised if the first total current flow (50) deviates (58) from the second total current flow (54) by more than a tolerance value (46).

Description

description

Method and device for monitoring a photovoltaic system

The invention relates to a method and a device for monitoring a photovoltaic system with a number of strings connected in parallel and guided against a common connection conductor to a return current.

A photovoltaic system (PV system) as a DC electrical system usually has a plurality of parallel strings connected in parallel, each of which in turn comprise a number of series-connected photovoltaic modules (PV modules). With the common connection conductors, against which the individual strings are guided, usually an inverter is connected. By means of this, the direct current provided by the PV modules is transformed into an alternating current for feeding into a power grid. If one or more of the PV modules are only slightly irradiated by solar energy compared to the others, or if individual PV modules are defective, it is possible that a so-called reverse current will occur. In this case, the electric current flows counter to the current direction of the fault-free case. As a result, individual P PV modules can be destroyed or the efficiency of the PV system is at least reduced.

From EP 2 282 388 A1 a device for feeding electrical energy from a photovoltaic system into a power grid is known. The photovoltaic system includes a plurality of strings with photovoltaic modules, each string can be switched on and off by means of a circuit breaker. For this purpose, the circuit breaker is opened or closed by a motor. In the case of an overcurrent, which is detected by means of a direction-sensitive current sensor associated with the respective string, the respective string is disconnected from the power supply, as is the case, for example, with faulty wiring. Furthermore, with each of the current sensors, the respective string is monitored for a return current, and in this case likewise the string is disconnected from the power supply. Due to the design of the current sensors, the directional positive current sensors, an assignment of the current flow in the respective direction due to the respective measurement sign possible.

The invention has for its object to provide a particularly suitable method and a particularly suitable device for monitoring an electrical PV system to a return current, which are particularly cost.

According to the invention the object with regard to the method by the features of claim 1 and with respect to the device by the features of claim 6. Advantageous developments and refinements are the subject of the respective subclaims.

The method is used to monitor a PV system for a return current, which comprises a number of strings connected in parallel and guided against a common connection conductor. A string indicates a rung. It is possible that at least one of the strings also consists of a further number of current paths connected in parallel to one another or at least has them. Conveniently, the system comprises two connection conductors, with which the strings are electrically connected in each case. Alternatively, the end of the string or strings facing away from the connecting conductor is in each case grounded. Suitably, the PV system comprises an inverter. The strings suitably each have a number of series-connected PV modules.

In each of the strings, a current flow is detected in a preferred direction, wherein the preferred direction is in particular rectified. In other words, the preferred direction of the individual strings is parallel to one another and either directed towards or away from the connection conductor. The detected current flows are preferably measured values which are detected by means of a suitable measuring device. Alternatively, the current flows are calculated from auxiliary measured variables. In the range of the respective tolerances contributing to the detection, the detected current flows correspond to the electrical currents actually flowing through the string. From the recorded current values, a summation current flow is created by, for example, adding up the individual values recorded. Appropriately, the value zero is used for the current flow in a current flow counter to the preferred direction. In a further step, which preferably takes place substantially at the same time, a current flow in the preferred direction in the connection conductor is detected and from this a second summation current flow is established, wherein the detected current flow is likewise in particular a directly measured value. In particular, the detected current flow through the connection conductor is used for the second summation current flow. Again, a current flow contrary to the preferred direction corresponds to a detected value of zero. In other words, the second summation current flow is equal to zero when the current flow in the connection conductor is opposite to the preferred direction. Preferably, all preferred directions are parallel to each other.

In a subsequent method step, the first summation current flow is compared with the second summation current flow. If the first summation current flow deviates from the second summation current flow by more than a tolerance value which is, for example, zero, a return flow is detected. In particular, there is a return flow if the second summation flow is less than the first summation flow plus the tolerance value.

By means of the method, the monitoring of the PV system is simplified, since only the current flow in one direction is monitored. An elaborate polarity reversal of measuring instruments, which could lead to an arc, is eliminated. Furthermore, only relatively inexpensive measuring devices can be used to determine the current flow, by means of which only a current flow in a certain direction can be determined, or which are not sensitive to direction.

Suitably, the preferred direction is selected counter to the respective reverse flow direction, ie it is antiparallel to this. Consequently, the respectively detected current flow does not correspond to the return current itself, but to the current in the desired current direction, for which the PV system is provided and in particular set up is. Thus, it is possible to determine the current power or the current efficiency of the PV system by means of the or each determined current flow, wherein the first or the second cumulative current flow is analyzed as the only method step.

Conveniently, that of the strings is determined as the carrier of the return current, which has the lowest determined current flow. In other words, it is assumed that the string has flowed through the return current at which the lowest current flow was determined. In particular, the deviation between the first and second sum current flow minus the tolerance value is used as the value of the return current flowing through the string with the lowest determined current flow. In this way, not only the fact is found that there is a backflow. It is also known, at least approximately, the value of the return current, even if this is associated with an error, namely the tolerance value itself. Thus, depending on the level of the return current and the associated tolerance value, the operation of the PV system can be adjusted or changed, in particular to prevent further propagation of the return flow.

In a particularly suitable embodiment, the string with the lowest determined current flow is separated from the connecting conductor, if a return current was detected. In this way, the string and any electrical components and / or further components of the PV system which are located in this current path are protected against further damage by the reverse current. Furthermore, a reduction in the efficiency of the PV system is prevented.

If, after disconnecting the string, there is still a return current which has now been detected by means of the summation current flows determined after the separation of the string, the string now having the lowest current flow is disconnected from the connecting conductor so that two strings are separated from the connecting conductor. Alternatively or in combination with this, the connection conductor itself is cut through in order to prevent the current flow through the connection conductor. If only the connection conductor is interrupted at a detected return current, the effort to prevent the backflow is comparatively low. If both the string and the terminal lead are disconnected, the safety against reverse current is increased because the power interruption is redundant. As an alternative to interrupting the connection conductor, it is also possible to interrupt all strings if a return current has been determined. Consequently, a current flow through the PV system is also prevented. This is provided, for example, if operation of the PV system with the number of strings reduced by one string is not possible or desired.

Conveniently, the sum of all valid error tolerances that prevail in detecting current flow through the strings is used to form the tolerance value. In particular, their sum forms the tolerance value. In other words, in addition to the current flow through the respective string, the prevailing respective fault tolerances are detected and added to form the tolerance value. It is possible that the individual error tolerances differ between the strings, and / or that depending on the level of the determined current flow different error tolerances are used.

Alternatively or particularly preferably in combination with this, the fault tolerance in the detection of the current flow through the connecting conductor is determined and this value is used to form the tolerance value. Expediently, the tolerance value is taken from the sum of the error tolerances that occur during the detection of the current flow through the respective strings, plus the fault tolerance in the detection of the current flow through the connection conductor for forming the tolerance value, and forms in particular these. If the error tolerances do not fluctuate symmetrically about the detected value, but are configured differently depending on the undershoot or violation, the sum of the positive error tolerances in the detection of the current flow through the strings plus the negative error tolerance in the detection of the tolerance value is formed Current flow through the connecting conductor or the sum of the negative error tolerances in the detection of the current flow through the strings plus the positive fault tolerance in the detection of the current flow through the connection conductor used as a tolerance value, depending on the sign of the deviation, that is, depending on whether the first sum current flow is greater or smaller than the second total current flow. In this case, "negative fault tolerance" denotes the deviation which is accepted in the determination of the respective current flow downwards, that is to say by how much the detected value deviates from the actual value In addition to the individual fault tolerances, for example, another correction term for forming the tolerance value Due to such a determination of the tolerance value, only actual return currents are recognized and not any artefacts determined on the basis of measurement accuracies In particular, if after a detected return current one or each current flow is interrupted, in this way the reliability is PV system increased.

The device for monitoring a PV system for a return current comprises a sensor arrangement and in particular a control unit, by means of which, for example, the method is carried out. In other words, the controller is provided and configured to perform the process of monitoring the PV system for reverse flow. The PV system has a number of strings (current paths) connected in parallel to one another and a connecting conductor against which, in particular all, strings are routed.

The sensor arrangement comprises a second current sensor and a number of first current sensors, that is to say at least two current sensors, which are provided and arranged to detect a current flow in each of the strings. For this purpose, each of the strings of the PV system is preferably assigned to one of the first current sensors in each case, and the number of first current sensors is expediently equal to the number of strings. By means of the second current sensor, a current flow through the connection conductor is detected during operation of the device.

The current detection takes place in the individual strings and the connecting conductor in a predetermined preferred direction. The preferred directions of the individual strings and the connection conductor are preferably parallel to each other and rectified and expediently contrary to the direction of the monitored return current.

The number of current sensors of the sensor arrangement suitably corresponds to the number of strings plus the connection conductor, which is the most cost-effective alternative, yet all current flows can be detected. The current sensors are designed, for example, as shunt resistors or as Hall sensors, which are arranged in an air gap of a slotted toroidal core, which is arranged around the respective current path, that is to say the connecting conductor or the respective string.

By means of the sensor arrangement, it is possible to detect not only the current flow of the strings but also the current flow through the connection conductor. This makes it possible to draw conclusions about a possibly flowing return current as well as a monitoring of the PV system on its performance, without the current flows of the individual strings needing to be added to this determination. Rather, this value is directly available, which also has a lower fault tolerance, since this is detected by means of a specially suitable current sensor and thus does not have the error tolerances of the first current sensors must be added. Compared to the use of only one current sensor monitoring the connection conductor, monitoring of the individual strings is also made possible by means of the use of the first current sensors.

Particularly preferably, the second current sensor is set up and provided only for detecting a current flow in the preferred direction. In other words, the second current sensor is not sensitive to direction. Consequently, a comparatively inexpensive current sensor and / or evaluation electronics associated therewith can be used. In the case of an actual electric current flowing in the direction opposite to the preferred direction, the detected current flow is therefore equal to zero (0).

Alternatively or particularly preferably, the first current sensors are not direction-sensitive, so that only a detection of the Current flow in the preferred direction is possible. If all the current sensors of the sensor arrangement are provided and set up only for detecting the current flow in the respective preferred direction, a comparatively inexpensive device can thus be realized.

Suitably, all first current sensors have the same fault tolerance. Consequently, such current sensors are selected and / or provided for the first current sensors, which are associated with the approximately same error tolerances. However, the individual measurement errors of the current sensors, ie the actual deviation in comparison to the manufacturer-specified fluctuation range of the measured values (fault tolerance), may differ between the individual current sensors.

By means of the use of current sensors with the same fault tolerance, an implementation of the method is particularly simplified, especially when the sum of all valid error tolerances of the first current sensors is used as a tolerance value or at least used to calculate the tolerance value. Particularly preferably, the individual first current sensors are identical to each other, which reduces maintenance.

Conveniently, moreover, the fault tolerance of the second current sensor is equal to the fault tolerance of the first current sensors, and in particular the second current sensor is identical to the first current sensors. Consequently, the sensor arrangement comprises only one type of current sensors, which are divided in the first current sensors and the second current sensor. In this way, a storage of current sensors is simplified and reduced maintenance. In addition, the calculation of the tolerance value is particularly simplified if it is calculated from the sum of the error tolerances for detecting the current flow through the connection conductor and the strings, and if only one current sensor is used per string. The tolerance value in this case is equal to the product of the valid fault tolerance of the current sensor type multiplied by the number of strings plus one. For example, the device comprises an interruption unit, by means of which the connecting conductor is separable, or by means of which at least one electric current is interrupted by the connecting conductor in the event of a return current. In addition, it is possible to switch off in the event of an overcurrent or overload of the PV system using the interrupt unit.

Conveniently, the device comprises at least one interrupt unit for interrupting an electrical current through one of the strings. Suitably, a number of interruption units corresponding to the number of strings is part of the device, it being possible for each of these to separate one of the strings from the connection conductor, or by means of which at least one electrical current through the respective string can be prevented.

The interruption unit or units are preferably part of the current sensors or at least indirectly electrically contacted with them, so that the current sensors provided with the interruption unit are exchangeable module by module, if they should be damaged.

The interruption unit has, in particular, a mechanical switch which, in the untripped state of the interruption unit, that is to say when a current flow through the interruption unit is possible, is closed and thus current-carrying. The mechanical switch, which is spring-loaded, for example, is preferably bridged by means of semiconductor electronics. The semiconductor electronics comprises an electrical switch, for example a transistor and in particular an IGBT. Furthermore, the semiconductor electronics on a control input, which is in particular connected to the mechanical switch. When the mechanical switch opens, that is to say when the current flow through the interruption unit is interrupted, the semiconductor electronics are switched in an electrically conducting manner on account of an arc forming in the region of the mechanical switch. For this purpose, the semiconductor electronics preferably have an energy store, which is charged as a result of the arc within the duration of the arc and by means of which the semiconductor electronics is operated. Due to the current conductivity of the semiconductor circuit in the case of an arc, a current path comparatively low-resistance to the arc is connected in parallel to the light bottom, which leads to a comparatively early extinction of the arc and thus a comparatively low load on the interruption unit. ¹

An embodiment of the invention will be explained in more detail with reference to a drawing. Show:

1 shows schematically a photovoltaic system with five parallel between two leads parallel strings and current sensors,

2 shows in a flow chart a method for monitoring the photovoltaic system to a return current, and

3 shows one of the current sensors with an interruption unit.

Corresponding parts are provided with the same reference numerals in all figures.

1 schematically shows a photovoltaic system (PV system) 2 with five strings 8 connected in parallel between a first connecting conductor 4 and a second connecting conductor 6. In other words, each of the strings 8 is guided on one side against the first terminal conductor 4 and on the opposite side against the terminal conductor 6 and contacted with these electrically. The two connecting conductors 4, 6 are in turn electrically connected to an inverter in order to feed an electrical power generated by means of the PV system 2 into a power grid. Each of the strings 8 has five photovoltaic modules (PV modules) 10, which in turn are connected in series in the respective string 8. By means of the PV modules 10, a direct current is provided when exposed to sunlight. In other words, during operation of the PV system 2 in each of the strings, an electric current flows in a preferred direction 12, which add up to an electric current through the connection conductors 4, 6. For monitoring the PV system 2 to a return current I _R against the preferred direction 12 in a reverse current direction 14, which occurs as an example in the third string 8, and to destruction of the PV modules 10 or at least to a reduction in the efficiency of the PV Appendix 2, a device 16 is provided which comprises a control unit 18 and a sensor arrangement 20. The sensor arrangement 20 in turn has five first current sensors 22, one of which is assigned to one of the strings 8, and by means of which a current flow ls is detected by the respective string 8 in the preferred direction 12. Furthermore, a second current sensor 24 is part of the sensor arrangement 20, wherein it is assigned to the first connection conductor 4. By means of the second current sensor 24, a current flow through the connecting conductor 4 in the preferred direction 12 is detected.

The current sensors 22, 24 are of identical construction, so that all the measured values recorded with them have the same spring tolerance. This is, for example, 0.05 ampere (A) and is symmetrical about the respective measured value, that is to say the respective detected current flow ls. The measured values can be determined only in the preferred direction 12 by means of the current sensors 22, 24. In the case of an actual electric current in the return current direction 14, ie at a return current I _R , the value detected by means of the current sensors 22, 24 is therefore equal to zero (0).

Each of the current sensors 22, 24 further includes an interrupt unit 26 (FIG. 3) and is connected to the control unit 18 via a line 28. In this case, the current flows ls detected by means of the current sensors 22, 24 are transmitted via the lines 28 to the control unit 18 and the respective interruption units 26 are acted upon by the control unit 18 by means of the lines 28 with control signals. If an overload occurs, or a shutdown of the PV system 2 is provided, the first terminal conductor 4 and / or the strings 8 are separated by means of the control unit 18 via lines 28 controlled interruption units 26 of the respective current sensors 22, 24, and respective actual flowing electric current is interrupted. FIG. 2 schematically shows a method 30 for the operation of the device 16 in a flow chart. After a start event 32, which takes place, for example, every 2 seconds, the current flow ls in each of the strings 8 is detected by means of the first current sensors 22 in a first detection step 34. With proper operation of the PV system 2 and certain irradiation conditions, this per string 8 is equal to 1 ampere (1A). In the third string 8, however, due to a fault in one of the PV modules 10, the reverse current IR occurs in the reverse current direction 14, which can not be detected by means of the first current sensor 22 assigned to the third string 8. Rather, the value communicated to the controller 18 performing the method 30 is 0A, although the current magnitude of the return current is IR = 0.5 amperes.

In a second detection step 36, which takes place substantially simultaneously with the first detection step 34, the current flow I _A in the first connection conductor 4, which is 3.5 A, is detected by means of the second current sensor 24. In a third and fourth detection step 38, 40, the error tolerances of the current sensors 22, 24, which are valid during the detection of the respective current flows ls, by the strings 8 and the first connection conductor 4, respectively, are determined. These are due to the design of the current sensors 22, 24 each 0.05A. In a first summarizing step 42, the error tolerances of the first current sensors 22 are added up. The fault tolerance of all the first current sensors 22 is thus equal to 0.25A. In a second summing step 46, this value is added to the fault tolerance of the second current sensor 24, thus forming a tolerance value 46, which is consequently 0.3A.

In a third summing step 48, all current flows ls assigned to the individual strings 8 are added up, which were determined in the first acquisition step 34, and thus a first summation current flow 50 is created. The first summation current flow 50 is therefore 4A. Furthermore, the value detected by means of the second current sensor 24 for the flow of current through the first connection conductor 4 in a fourth summation step 52 is used as the second summation current flow 54. The second total current flow 54 is therefore 3.5A. In a comparison step 56, a deviation 58 of the two summation current rivers created by each other. The deviation 58 used here is the amount of the difference between the two. In other words, the deviation 58 is 0.5A. The deviation 58 is compared with the tolerance value 46, which is 0.3A. If the deviation 58 is less than the tolerance value 46, the method 30 is ended in a first end event 60.

If the deviation 58 is greater than the tolerance value 46, the string 8 which carries the return current IR is determined in a determination step 62 on the one hand. This is the third string whose value for the detected current flow Is in the preferred direction 12 is lowest, namely zero (0). Furthermore, the difference between the two is taken as the value for the return current I _R , ie 0.2A. In an interrupting step 64, the interruption unit 26 of the first current sensor 22 of the third string 8 disconnects it from the first connecting conductor 4, and thus the reverse current I _{R is} interrupted. Consequently, the current flow through the first lead 4 increases from 3.5A to 4A. After separating the third string 8 from the first lead 4, a second end event 66 occurs and the method 30 is terminated.

3 shows a comparatively detailed circuit diagram of one of the suitable current sensors 22, 24 with the interruption unit 26. The current sensor 22, 24 comprises a main current path 68 with a measuring sensor 70 which is connected in the main current path 68 and which measures the current flow ls in which Preferred direction 12 detected. The main current path 68 leads through the interruption unit 26, which has a switching contact 72, also referred to below as a mechanical switch, and a semiconductor electronics 74 connected in parallel therewith. The mechanical switch 72 and the semiconductor electronics 74 form a self-sufficient hybrid disconnect switch.

The semiconductor electronics 74 essentially comprises two semiconductor switches 73a, 73b, which are connected in parallel to the mechanical switch 72, and a drive circuit 76 with an energy store 78 and with a timer 80. The drive circuit 76 is, preferably via a resistor or a resistor row R, with connected to the main current path 68. The gate of a preferential wise IGBT used as a semiconductor switch 73a, 73b forms the control input 82 of the semiconductor circuit 74. This control input 82 is guided via the drive circuit 76 to the main current path 68.

The first semiconductor switch (IGBT) 73a is connected in a cascode arrangement with the second semiconductor switch 73b in the form of a MOSFET in series. The potential U + applied to the first semiconductor switch 73a is always greater than the potential U- on the opposite side of the switch, at which the second semiconductor switch (MOSFET) 73b is led to the main circuit 6. The positive potential U ₊ is 0V when the mechanical switch 72 is closed.

The first semiconductor switch (IGBT) 73a is connected to a freewheeling diode D2. A first Zener diode D3 is connected on the anode side to the potential U. and on the cathode side to the gate (control input 82) of the first semiconductor switch (IGBT) 73a. A further Zener diode D4 is in turn connected to the gate (control input 82) on the cathode side and to the emitter of the first semiconductor switch (IGBT) 73a on the anode side.

At a center or Kaskodenabgriff 84 between the first and second semiconductor switches 73a and 73b of the cascode arrangement on the anode side, a diode D1 is performed, which is connected on the cathode side via a serving as energy storage capacitor 78 C against the potential U. Also, several capacitors C can form the energy storage 78. Via an anode-side voltage tap 86 between the diode D1 and the energy store 78 or the capacitor C, a transistor T1 connected to ohmic resistors R1 and R2 is connected via further resistors R3 and R4 to the gate of the second semiconductor switch, which in turn is led to the control input 82 of the semiconductor electronics 74 (MOSFET) 82 connected. Another Zener diode D5 with parallel resistor R5 is connected on the cathode side to the gate and on the anode side to the emitter of the second semiconductor switch (MOSFET) 73b.

On the base side of the transistor T1 is driven by a transistor T2, in turn, the base side via an ohmic resistor R6 with the example connected as Monoflopp running timer 80 is connected. Base-emitter side, the transistor T2 is also connected to a further resistor R7.

When the mechanical switch 72 is closed, the main current path 68 is low-ohmic, while the parallel commutation path 88 of the hybrid disconnecting switch 72, 74 formed by the semiconductor switches 73a, 73b is high-impedance and thus current-blocking. Before the opening of the mechanical switch 72, the electrical voltage occurring there is practically 0V and increases suddenly with the opening of the switch contacts 72a, 72b of the mechanical switch 72 to a value characteristic of an arc LB having a typical arc voltage ULB of, for example, 20V to 30V. The positive potential U + thus goes against this arc voltage U _L B ~ 30V when the mechanical switch 72 opens.

During the period of time following the contact opening time (arc time interval), the commutation of the switch current Is,, I _R, which corresponds essentially to the arc current, already starts from the main current path 68 to the commutation path 88.

During the arc time interval practically the arc current l _s,. I _R between the main current path 68 - ie via the mechanical switch 72 - and the commutation path 88 - so the semiconductor electronics 74 on. During this arc time interval, the energy store 78 is charged. The time duration is set such that, on the one hand, sufficient energy is available for reliably triggering the semiconductor electronics 74, in particular for switching them off during a period following the arc time interval. On the other hand, the arc time interval is sufficiently short that undesirable contact erosion or wear of the switch 72 or the switch contacts 72a, 72b is avoided.

With the onset of the arc LB and thus when the arc voltage is generated, the first semiconductor switch (IGBT) 73a is at least as far controlled through the resistor R, that a sufficient charging voltage and a sufficient reaching arc or charging current for the capacitors C and thus for the energy storage 78 is available. For this purpose, with the corresponding circuit of the first semiconductor switch (IGBT) 73a with the resistor R and the diode D3, a control circuit of the electronics 74 is preferably provided, with which the voltage at the cascode tap 84 is set to, for example, U _Ab = 12V (DC). In this case, a fraction of the arc current and thus of the switch current ls,, IR flows through the positive potential U + near the first semiconductor switch (IGBT) 73a of the hybrid circuit breaker 72,74.

The resulting pick-off voltage serves to supply the drive circuit 76 of the electronics 74 formed essentially by the transistors T1 and T2 as well as the timer 80 and the energy store 78. The diode D1 connected to the cascode tap 84 on the cathode side and the capacitor C on the cathode side prevents the charging current from flowing back from the capacitors C and via the commutation path 88 in the direction of the potential IL. If sufficient energy is contained in the capacitor C and thus in the energy store 78, and consequently a sufficiently high control or switching voltage is present at the voltage tap 86, then the transistor T1 and consequently the transistor T2 control, so that the two semiconductor switches 73a, 73b completely go through. The arc or switch current ls, IA. IR flows due to the compared to the very high resistance of the formed by the open switch 72 separation distance of the main current path 68 substantially lower resistance of the now durchgesteuerte semiconductor switches 73a, 73b almost exclusively on the Kommutierungspfad 88. The positive potential U + is thus again against 0V when the switch current l _s,. IR commutated to the electronics 74. As a result, the arc LB between the contacts 72a, 72b of the mechanical switch 72 extinguished.

The charge capacitance and thus the storage energy contained in the capacitor C is dimensioned such that the semiconductor electronics 74 carries the switch current ls,, l _R for a time period predetermined by the timer 80. This period of time can be set to, for example, 3 ms. The dimensioning of this time duration and thus the definition of the timer 80 depends essentially on the application-specific or typical time periods for a complete extinction of the arc LB and after a sufficient cooling of the plasma formed. The essential proviso here is that after switching off the electronics 74 with high-impedance commutation path 88 and consequently current-blocking semiconductor electronics 74 at the still open mechanical switch 72 or via its switch contacts 72a, 72b no renewed arc LB can arise.

After the expiration of the time period set by the timer 80, the switch current ls,, IR decreases to practically zero (ls,, IR = 0A), while at the same time the switch voltage rises to the operating voltage provided by the strings 8. The positive potential U + thus goes against this operating voltage when the commutation path 88 becomes high-coherent as a result of the blocking of the semiconductor switches 73a, 73b and thus the electronics 74 again becomes current-blocking.

If a direct-current (DC) system has at least two current paths guided to one another and against a common potential point, the method can also be advantageously applied to such an electrical system.

LIST OF REFERENCE NUMBERS

2 photovoltaic system

4 first connection conductor

6 second connection conductor

8 string

10 photovoltaic module

12 preferred direction

14 reverse current direction

16 device

18 control unit

20 sensor arrangement

22 first current sensor

24 second current sensor

26 interruption unit

28 line

30 procedures

32 start event

34 first detection step

36 second detection step 8 third detection step 0 fourth detection step 2 first summing step 4 second summing step 6 tolerance value

8 third summary step 0 first summation current flow 2 fourth summation step 4 second summation current flow 6 comparison step

8 deviation

0 first end event

2 determination step 64 interruption step

66 second end event

68 main current path

70 measuring sensor

72 switches

72a, b switch contacts

73a first semiconductor switch

73b second semiconductor switch

74 semiconductor electronics

76 drive circuit

78 energy storage

80 timer

82 control input

84 Middle or cascade tap

86 voltage tap

88 commutation path

C capacitor

D1 diode

D2 freewheeling diode

D3 - D5 Zener diode

l _s current flow string

l _A current flow connecting conductor l _R return current

LB arc

R, R1 - R7 ohmic resistance

T1, T2 transistor

U ₊ , U. potential

Claims

claims

1. A method (30) for monitoring a photovoltaic system (2) with a number of strings connected in parallel and against a common connection conductor (4, 6) (8) to a return current (l _R ), in which

- detects a current flow (Is) in a preferred direction (12) in each of the strings (8) and from this a first summation current flow (50) is created,

- detects a current flow () in the preferred direction (12) in the connecting conductor (4, 6) and from this a second summation current flow (54) is created,

- the first summation flow (50) is compared with the second summation flow (54), and

- When a deviation (58) of the first summation current flow (50) from the second summation current flow (54) by more than a tolerance value (46), a return current (l _R ) is detected.

2. Method (30) according to claim 1,

characterized,

the preferred direction (12) is selected counter to the respective return flow direction (14).

3. Method (30) according to claim 1 or 2,

characterized,

that the string (8) with the lowest determined current flow (Is) is determined as the carrier of the return current (l _R ), and in particular the deviation (58) minus the tolerance value (46) is used as the value of the return current (l _R ).

4. Method (30) according to one of claims 1 to 3,

characterized,

in the case of a detected reverse current (IR), the string (8) having the lowest determined current flow (Is) is disconnected from the connecting conductor (4, 6) and / or the connecting conductor (4, 6) itself.

5. Method (30) according to one of claims 1 to 4,

characterized,

that the sum of all valid error tolerances in the detection of the current flow (Is) by the strings (8) and / or the applicable fault tolerance in the detection of the current flow (l _A ) through the terminal conductor (4, 6) is used as a tolerance value (46) ,

6. Device (16) for monitoring a photovoltaic system (2) with a number of parallel connected and against a common connection conductor (4, 6) guided strings (8) to a return current (l _R ), in particular for carrying out the method (30) according to one of claims 1 to 5, with a sensor arrangement (20) having a, in particular to the number of strings (8) corresponding number of first current sensors (22) and a second current sensor (24) for detecting a current flow (Is, IA ) by each string (8) or by the connecting conductor (4, 6) in a preferred direction (12).

7. Device (16) according to claim 6,

characterized,

the current sensor (s) (22, 24) is / are provided and set up only for measurement in the preferred direction (12).

8. Device (16) according to claim 6 or 7,

characterized,

in that all the first current sensors (22) have the same fault tolerance, which in particular is equal to a fault tolerance of the second current sensor (24).

9. Device (16) according to one of claims 6 to 8,

marked by

an interruption unit (26) for interrupting the current flow (ls, l _A ) through at least one of the strings (8) and / or through the connection conductor (4, 6).

10. Device (16) according to claim 9,

characterized,

in that the interruption unit (26) comprises a current-carrying mechanical switch (72) and semiconductor electronics (74) connected in parallel therewith, which is current-blocking when the switch (72) is closed, and which has a control input (82) connected to the switch (72) in such a way in that, when the switch (72) opens, an arc voltage generated as a result of an arc (LB) across the switch causes the semiconductor electronics (74) to conduct current.