WO2025214646A1 - Circuit breaker - Google Patents

Abstract

The invention relates to the protection of a power supply formed by three phase conductors (L1, L2, L3) and a neutral conductor (N), said protection being ensured by means of a circuit breaker. The circuit breaker is designed to determine current values of the three phase conductors (L1, L2, L3), determine values of the voltages applied between the phase conductors (L1, L2, L3) and the neutral conductor (N), evaluate determined current and voltage values with respect to the fulfillment of conditions for switching actions which relate to the interruption of the phase conductors (L1, L2, L3) and the restoration of interrupted connections, output control signals in accordance with switching actions to be carried out according to the evaluation results and, as a result of the control signals, interrupt the phase conductors (L1, L2, L3) and restore interrupted connections. Furthermore, for at least one switching action to be carried out, the circuit breaker is designed to use a criterion for a possible overcurrent when carrying out said switching action, determine, from the determined current values, whether this criterion has been met and carry out the switching action in accordance with the criterion. The invention allows a flexible protection for a power supply formed by three phase conductors and a neutral conductor.

Description

Description

circuit breaker

The invention relates to a circuit breaker and a method for protecting a power supply formed by three phase conductors and one neutral conductor.

Loads with capacitive or inductive components (inrush loads) generate high inrush currents when connected to an AC network. For example, a load consisting of one or more single- or multi-phase rectifiers or switching power supplies can generate brief, high current peaks (inrush currents) when connected, the maximum values of which can reach several times their steady-state current amplitude. High-efficiency electrical machines also have inrush currents and starting currents that are significantly greater than their rated currents, e.g., up to a factor of 12 for machines in efficiency class IE4.

One consequence of the inrush current can be the tripping of the miniature circuit breaker (MCB) when an inrush load is switched on or connected. For two-pole inrush loads, the amplitude of the inrush current and thus the probability of tripping due to an inrush current is determined by various factors. The most important of these are:

• the load impedance

• the time at which this circuit is connected to the current-driving voltage

• the state of the load immediately before the switching-on time (e.g. for capacitive loads the charge of the capacitance and for electrical machines the amplitude and phase position of the induced voltage)

For a capacitive load or a rectifier with a capacitive intermediate circuit, the tripping probability is lowest when the load is connected near the zero crossing of the relevant voltage. This is because, at such a selected switch-on time, the current rise rate is generally the lowest compared to other switch-on times. Furthermore, the time in the switched-on state until the current limit is reached is the longest (in case excessive current values are reached during an in-rush current). This allows a greater amount of energy or charge to be transferred to the load than with a very short and steep current pulse. Furthermore, more time is available for analyzing the load in the switched-on state. With conventional electromechanical circuit breakers with manual contact actuation, there is no way to specifically implement an optimal switch-on time. In contrast, with switching devices with an electronic switching unit (Semiconductor Circuit Breaker (SCCB)) consisting of one or more power semiconductors, one or more energy absorbers (EA), a measuring unit, and a control unit, targeted switching at a selected time with a tolerance of a few microseconds is possible.

German Patent Application DE 102020 216405 A1 discloses a method for controlling the power semiconductor in a two-pole SCCB. The inrush current is switched off when a certain current limit is reached and then repeatedly switched on within a mains voltage period near the next voltage zero crossing. During the first switch-on, the capacitance of the inrush load may be partially charged, and the amplitude of the inrush current decreases with each subsequent switch-on attempt, so that the current limit of the SCCB may not be reached again during one of the subsequent switch-on attempts. After that, the load is continuously supplied.

In three-phase four-wire systems, the problem of inrush currents is more complex due to the higher number of conductors and requires different solutions than the method mentioned above. The invention aims to contribute to the protection of a power supply.

The object is achieved by a circuit breaker according to claim 1 and a method according to claim 13. Advantageous embodiments are specified in the subclaims.

The circuit breaker according to the invention is designed to protect a power supply consisting of three phase conductors and one neutral conductor. For this purpose, it is configured to determine the current values of the three phase conductors and the voltage values between the phase conductors and the neutral conductor. This determination is typically performed by measurement, for which purpose it can be configured with current and voltage sensors. However, it is also conceivable that these values are calculated from other measured values. Solutions are also conceivable in which the measuring sensors are not part of the circuit breaker, but are measured externally and transmitted to the circuit breaker. In this case, the term "determine" is to be understood as receiving or querying an external entity. The circuit breaker according to the invention is further designed to evaluate specific current and voltage values with a view to fulfilling conditions for switching operations involving the interruption of the phase conductors and the restoration of interrupted connections. For this purpose, it can be equipped with one or more processing units, e.g., an MCU or CPU.

Furthermore, the circuit breaker according to the invention is designed to output control signals in accordance with switching operations to be performed based on the evaluation results, and to interrupt the phase conductors and restore broken connections triggered by the control signals. A "switching operation" generally comprises at least one condition and one action, e.g., an overcurrent on a phase conductor and an interruption of the phase conductor. It is also possible that several conditions must be met for the action to occur.

The circuit breaker is preferably equipped with switching transistors (e.g., MOSFETs or IGBTs) to interrupt the phase conductors. Galvanic isolation can also be provided. A corresponding switch with a phase conductor and a neutral conductor is described, for example, in DE 102022201960 A1. The term "circuit breaker" is to be understood as meaning that the switch has protective functions (e.g., overcurrent protection). The switch can also perform other functions unrelated to protection, such as operational switching.

According to the invention, the circuit breaker is designed to be able to carry out at least one switching operation

- to use a criterion for a possible overcurrent when carrying out this switching operation,

- to determine from the determined current values whether this criterion is met, and

- to carry out the switching operation in accordance with this criterion.

Typically, different criteria are provided depending on the switching operation. The term "possible overcurrent" refers to a situation that can occur with the switching operation or for which there is a significant probability of it occurring. A corresponding assessment can be made, for example, based on a previous state of the circuit breaker, e.g. the occurrence of an overcurrent situation during a previous interruption of one or more phases. This means that the criterion can be based on an assessment or evaluation of previously determined current or voltage values. However, it is also It is conceivable that additional variables determined by sensors, e.g. temperature information, are included.

According to one embodiment, a switching operation comprises restoring at least one interrupted connection of a phase conductor, and the circuit breaker is configured not to perform the switching operation if the criterion is met. That is, the switching operation is omitted because there is a significant probability of an overcurrent event occurring after the switching operation has been performed.

According to one embodiment of the circuit breaker, a switching operation involves restoring a broken connection of a phase conductor. Furthermore, a virtual circuit is defined for the phase conductor and the neutral conductor, and the circuit breaker is configured to not perform the switching operation if the criterion for a potential overcurrent is met on this virtual circuit. The term "virtual circuit" is to be understood as simulating that the two affected conductors (here, a phase conductor and the neutral conductor) form a circuit, and on this basis, values are determined for this virtual circuit, which can then be checked for compliance with limit values.

According to one embodiment, a switching operation of the circuit breaker comprises restoring the interrupted connections of two phase conductors, and two virtual circuits are defined, one for each of the phase conductors and the neutral conductor. In this embodiment, the circuit breaker is configured not to perform the switching operation if the criterion for a possible overcurrent is met for one of the two virtual circuits.

According to one embodiment, a switching operation relates to the interruption of at least one phase conductor, and the circuit breaker is configured to carry out the switching operation when the criterion is met. According to a further development of this embodiment, a virtual circuit with another phase conductor is defined for the phase conductor, and the criterion for possible overcurrent is met for this virtual circuit, wherein the other phase conductor is switched off but fulfills a sufficient condition for restoring the connection. The interruption of the phase conductor is then carried out with a view to restoring the connection of another phase conductor, wherein an overcurrent is avoided during this restoration on the virtual circuit of the two phase conductors. According to one embodiment of the circuit breaker according to the invention, a switching action to be performed according to a criterion for possible overcurrent is assigned to a virtual circuit. The circuit breaker is then configured to

- to calculate the current values characterising the virtual circuit from the measured current values and

- to consider the criterion for a possible overcurrent as fulfilled if i) a phase conductor of the virtual circuit was interrupted due to an overcurrent event, and ii) the current values characterising the virtual circuit immediately before the interruption of the phase conductor fulfilled at least one condition for overcurrent. This can be a virtual circuit with a phase conductor and the neutral conductor, and the at least one condition consists in the absolute value of the weighted sum of the current values of the three phases (“neutral conductor current”) being greater than the absolute values of the weighted differences between the current values of the phase conductor and the current values of the other two phase conductors. However, it can also be a virtual circuit with a first

phase conductor and a second phase conductor, wherein the first phase conductor was interrupted due to an overcurrent event, and the at least one condition is that the amount of the weighted difference between the current values of the phase conductors of the virtual circuit is greater than the amount of the weighted sum of the current values of the three phase conductors (“neutral conductor current”) and greater than the amount of the weighted difference between the current values of the first phase conductor and the third phase conductor not belonging to the virtual circuit.

According to one embodiment of the invention, a switching operation to be performed according to a criterion for a possible overcurrent is assigned to a virtual circuit, and the circuit breaker is configured to set the criterion for a possible overcurrent as not met for this virtual circuit if all phase conductors (one or two) of the virtual circuit are conducting without overcurrent. This can be implemented, for example, by means of a flag that is set after a corresponding overcurrent event and cleared again after conducting without overcurrent.

The term "overcurrent-free conducting" refers to stable operation of phase conductors. A condition can be defined for this, e.g., by means of a time threshold for the phase conductors of the virtual circuit to conduct without interruption after an overcurrent event. According to one embodiment of the circuit breaker according to the invention, various switching actions are defined for the circuit breaker, which comprise reconnecting at least one phase conductor, and the circuit breaker is designed to carry out only a subset of these switching actions in accordance with a criterion for possible overcurrent. It is expedient to carry out at least one switching action of the plurality of switching actions independently of criteria for a possible overcurrent, i.e. the plurality of switching actions is divided into a group that is carried out independently of a criterion for possible overcurrent, and a group in which the execution of the switching action depends on the criterion. The criterion can, for example, relate to an overcurrent situation before the phase conductors are interrupted. If all switching actions depend on such a criterion, it may happen that the circuit breaker can no longer switch on. Therefore, a group of switching operations is preferably defined that do not depend on such a criterion and allow the restoration of the connections of all three phase conductors, e.g., switching on when the magnitude of the phase conductor voltage is close to zero (as opposed to switching depending on a voltage difference between two phase conductors). For example, switching operations can be defined that provide for the restoration of a connection for a phase conductor based on the voltage between the phase conductor and the neutral conductor, and these switching operations are not performed based on a criterion for possible overcurrent.

The invention also relates to a method for protecting a power supply formed by three phase conductors and a neutral conductor, which method is preferably carried out by means of a circuit breaker according to the invention.

The invention is described in more detail below using exemplary embodiments.

Fig. 1: a schematic representation of the general problem with an abstract protective switching device,

Fig. 2: Voltages in a symmetrical four-wire system,

Fig. 3: an SCCB for a four-wire three-phase system, Fig. 4: a switch-on without overcurrent events starting with switching operation SH1 (scenario 1),

Fig. 5: a switch-on without overcurrent events starting with switching operation SH2 (scenario 2),

Fig. 6: a switch-on with a short circuit in phase 3 starting with switching operation SH1 (scenario 3),

Fig. 7: a switch-on with a short circuit in phase 2 (scenario 4),

Fig. 8: a switch-on with a short circuit between phases 2 and 3 (scenario 5),

Fig. 9: an execution of switching operation SH 5 (scenario 6),

Fig. 10: a switch-on with two single-phase short-circuits in phases 3 and 1 using the basic switching procedure (scenario 7),

Fig. 11 : a switch-on with two single-phase short-circuits in phases 3 and 1 with the basic switching procedure and additional conditions ZB1 and ZB2 (scenario 8),

Fig. 12: a switch-on with a single-phase short circuit in phase 3 with the

Basic switching procedure and additional conditions ZB1 and ZB2 (Scenario 9),

Fig. 13: a switch-on with a single-phase short circuit in phase 3 using the basic switching method and

Additional conditions ZB1, ZB2 and ZB3 (Scenario 10),

Fig. 14: a switch-on with a two-phase short circuit between phases 2 and 3 with the basic switching procedure and additional conditions ZB1, ZB2 and ZB3 (scenario 11),

Fig. 15: a switch-on with a two-phase short circuit between phases 2 and 3 with the basic switching procedure and additional conditions ZB1, ZB2 and ZB3 and the switching operation SH7 (scenario 12), Fig. 16: a switch-on with a two-phase short circuit between phases 2 and 3 with the basic switching procedure and additional conditions ZB1, ZB2 and ZB3 as well as switching operations SH7 and SH8,

Fig. 17: a switch-on with a two-phase short-circuit between phases 2 and 3 with the basic switching procedure and additional conditions ZB1, ZB2 and ZB3 as well as switching operations SH7, SH8 and SH9 (scenario 14),

Fig. 18: a switch-on with a two-phase short-circuit between phases 2 and 3 with the basic switching procedure and additional conditions ZB1, ZB2 and ZB3 as well as switching operations SH7, SH8 and SH9 (scenario 15),

Fig. 19: switching on with the basic switching procedure with additional conditions ZB1 to ZB3 and switching operations SH7 and SH8 (scenario 16),

Fig. 20: a flowchart for a first method for checking for a self-switching off of a switch according to the invention (permanent switching off),

Fig. 21 : a flow chart for a second method for checking a self-switching off of a switch according to the invention (permanent switching off) and

Fig. 22: a flowchart regarding a reset of self-shutdown conditions according to Fig. 20 or Fig. 21.

Any load or load combination can be connected to a four-wire network with three phase conductors and one neutral conductor. This is shown in Fig. 1.

Between each of the phase conductors L1, L2, L3 and the neutral conductor N, a circuit with the impedances ZL1N, ZL2N, ZL3N can be formed, in which the currents iL1N, iL2N, iL3N are influenced by the phase voltages (voltages between the phase conductors and the neutral conductor) uL1N, uL2N, uL3N. These circuits are also referred to below as SKL1N, SKL2N and SKL3N. In addition, three independent circuits with the impedances ZL1L2, ZL2L3, ZL3L1 can be present between the phase conductors, in which the line-to-line voltages (voltages between two phase conductors) uL1L2, uL2L3, uL3L1 are decisive for the steepness and amplitude of the currents iL1L2, iL2L3, iL3L1. These circuits are also referred to as SKL1L2, SKL2L3 and SKL3L1. This results in a total of six circuits, each of which can contain any load, possibly including inrush loads or short circuits. 63 (2**6-1) combinations are possible, in which at least one current-conducting two-terminal device is present in one to six circuits.

Depending on the charge state of the intermediate circuit capacitance, single-phase and/or three-phase rectifiers may or may not be recognizable as loads at different times. Due to its mode of operation, a three-phase rectifier can conduct one or two phase-to-phase currents at different times and thus be recognizable as one or two line-to-line loads. The phase and amplitude of the currents i L1 N, i L2 N , i L3 N as well as i L1 L2, i L2 L3, i L3 L1 through the impedances Z L1 N, Z L2 N, Z L3 N as well as Z L1 L2, Z L2 L3, Z L3 L1 are determined accordingly by the voltages u L1 N, u L2 N, u L3 N as well as u L1 L2, u L2 L3, u L3 L1. In a symmetrical three-phase voltage system, the zero crossings of the three phase voltages uL1N, uL2N, and uL3N are offset from each other by one-sixth of the grid period, or 60 degrees (see Fig. 2). The phase positions of the phase-to-phase voltages uL1L2, uL2L3, and uL3L1 are also offset from each other by 60 degrees. In addition, each zero crossing of a phase-to-phase voltage is offset from a zero crossing of a phase voltage by one-twelfth of the grid period, or 30 degrees.

Conventional multi-pole circuit breakers only allow the simultaneous switching on and off of all phase switches SL1, SL2, SL3, and possibly also SN (see Fig. 1). Due to the voltage phase positions described above in a four-wire system, several of the voltages uL1 N, uL2N, uL3N, uL1 L2, uL2L3, and uL3L1 typically have values at any given time that can cause very high inrush currents. The probability of unwanted MCB tripping due to an inrush current when connecting the inrush load at one of the impedances ZL1N, ZL2N, ZL3N, ZL1 L2, ZL2L3, and ZL3L1 is not insignificant.

If the time-current characteristic of a circuit breaker is exceeded in one phase of a conventional MCB, this will trigger and permanently disconnect all protected conductors from the grid. The load can only be reconnected to the grid by manually reconnecting it.

An SCCB for a four-wire network is shown in Fig. 3. An SCCB with self-commutated

Power semiconductors, energy absorbers, a measuring and control unit as well as the The control unit with triggerable isolating contacts has the option of actively switching the phase semiconductors on and off at any time.

The duration and maximum value of the inrush current play a major role in the dimensioning of the power semiconductors and energy absorbers of an SCCB. The duration of an unlimitated inrush current can range from a few milliseconds to several grid voltage periods. Sizing the power semiconductor to carry the inrush current over such time intervals increases the cost and volume of these components. Limiting the inrush currents in a multi-pole SCCB is therefore of great importance when designing the switching and operating strategy.

Furthermore, the objective is to reduce and, if possible, minimize the probability of tripping due to inrush current when operating as many load combinations as possible on the multi-pole SCCB. A four-wire system must transition to steady-state as quickly as possible when the SCCB is switched on. When connecting a new inrush load, e.g., by closing one of the switches SL1N, SL2N, SL3N, SL1L2, SL2L3, SL3L1 in Fig. 1, the effect of the disturbance caused by the inrush current on the other loads must be minimal. In both cases, the inrush currents must be limited to a specific maximum value.

An SSCB as shown in Fig. 3 is assumed. Operating procedures specifically developed for such SSCBs are described below. Based on this, decisions are made to switch the power semiconductors of the individual phases on and off based on the detected overcurrent events, the current switching states of the power semiconductors, and the measured voltage and current values. In addition, the load side is continuously monitored in order to identify the likely causes of the overcurrent after an overcurrent event and, if necessary, to improve the quality of the switching decisions based on this. For the sake of simplicity, the following refers to the power semiconductor of one phase, although in reality, the shutdown can be realized using multiple electronic components or switching elements (in Fig. 3, anti-parallel transistors, each with a parallel protection diode). This scenario is intended to be encompassed by reference to a switched power semiconductor; specifically, the term "power semiconductor", when referring to a phase, also includes all switching devices formed by power switches. First, the basic switching method is described, which also defines the hardware requirements. The instantaneous values of the phase voltages uL1N, uL2N, and uL3N are recorded by the control unit. By measuring the three phase voltages, the voltage information from the four-wire system is fully captured—in contrast to the measurement of the phase-to-phase voltages, which omits the voltage to the neutral conductor. This is an important feature of the SCCB arrangement, which allows the basic switching method to be implemented.

The instantaneous values of the phase currents i 1 , i 2 , i 3 are recorded by the control unit. The phase-to-phase voltages uL1 L2, uL2 L3, uL3 L1 are calculated from the measured phase voltages uL1 N, uL2 N, uL3 N, thus eliminating the need for additional voltage measuring elements. The calculation rule for the phase-to-phase voltages is: o uL1 L2 = uL1 N - uL2 N o uL2 L3 = uL2 N - uL3 N o uL3 L1 = uL3 N - uL1 N

Another characteristic of the SSCB is phase-by-phase switching, i.e., the power semiconductor of each individual phase can be switched on or off at any time by the control unit, independently of the power semiconductors of the other phases. If a freely selectable current limit iMAX is exceeded in a phase (overcurrent event), the power semiconductor of that phase is switched off immediately. This current limit is not necessarily the same for all phases. The time from exceeding the current limit until the overcurrent is switched off is so short that the power semiconductor is not damaged by the brief overcurrent. The overcurrent event in one phase does not trigger automatic shutdown of the power semiconductors in the other phases.

The mechanical isolating contacts are controlled by the control unit independently of the power semiconductors and are not automatically triggered in the event of an overcurrent event. The switching states of the power semiconductors and the overcurrent events that occur are recorded by the control unit.

The following switching operations are provided or defined:

As soon as and as long as at least one power semiconductor is switched on: • Switching action SHO: when a current limit iMAX is exceeded in a phase conductor with a switched-on power semiconductor, switch off the power semiconductor of this phase.

Switching operations when power semiconductors of all phases are switched off (switching state SZO):

• Switching operation SH1: if the magnitude of the phase voltage between a phase conductor and the neutral conductor is less than a certain value uMAXLN(on), switch on the power semiconductor of this phase.

• Switching operation SH2: if the magnitude of the line-to-line voltage between two phase conductors is less than a certain value uMAXLL(on), switch on the power semiconductors of these two phases.

Switching operations when only the power semiconductor of one phase is switched on (switching state SZ1):

• Switching operation SH3: if the magnitude of the voltage between a phase conductor with the power semiconductor switched off and the neutral conductor is less than a certain value uMAXLN(on), switch on the power semiconductor of this phase.

• Switching operation SH4: if the magnitude of the line-to-line voltage between a phase conductor with the power semiconductor switched off and the phase conductor with the power semiconductor switched on is less than a certain value uMAXLL(on), switch on the power semiconductor of this switched off phase.

Switching operations when power semiconductors are switched on in only two phases (switching state SZ2):

• Switching operation SH5: if the voltage between the phase conductor with the power semiconductor switched off and the neutral conductor is less than a certain value uMAXLN(on), switch on the power semiconductor of this phase.

• Switching operation SH6: if the magnitude of the line-to-line voltage between the phase conductor with the power semiconductor switched off and a phase conductor with the power semiconductor switched on is less than a certain value uMAXLL(on), switch on the power semiconductor of the switched-off phase.

If power semiconductors are switched on in all three phase conductors (switching state SZ3), no switching operations are performed with the exception of switching operation SHO. The values uMAXLN(on), uMAXLL(on) should be set close to zero or in the range of 0...20 percent of the phase voltage amplitude.

The switching operations SHO, SH1, SH2, and SH4 are the basic switching operations. The others are derived from them and simply describe a different initial situation. Nevertheless, they are defined separately in order to describe and explain the process in detail within this application.

Switching operations SH3 and SH5 are generally identical to SH1. If at least one power semiconductor is off, it will be switched on when the phase voltage between its phase and neutral conductors is close to zero crossing. The only difference between switching operations SH1, SH3, and SH5 is the number of switched-on power semiconductors before the switching operation.

Switching operation SH6 is generally identical to switching operation SH4. If at least one power semiconductor is off, it will be switched on when the phase-to-phase voltage between its phase and another phase is close to zero. The only difference between switching operations SH4 and SH6 is the number of switched-on power semiconductors before the switching operation.

Based on the expected overcurrent events, the switching operations can be improved. For the following description, two circuit states are defined: each of the circuits SKL1 N, SKL2N, SKL3N, SKL1 L2, SKL2L3, SKL3L1 has the state

• “normal” if the probability of an overcurrent event is assessed as low when the respective circuit is connected to the grid.

• “Overcurrent” if the probability of an overcurrent event is assessed as high when the respective circuit is connected to the grid.

The following additional conditions for the execution of switching operations are defined:

• Additional condition ZB1 for the switching operation SH2: Carry out the switching operation SH2 only if the states of the circuits between each of the two phase conductors to be switched on according to the switching operation SH2 and the neutral conductor are “normal”.

• Additional condition ZB2 for switching operation SH4: Only carry out switching operation SH4 if the state of the circuit between the device to be switched on according to switching operation SH4 phase conductor and the neutral conductor is “normal”.

• Additional condition ZB3 for switching operation SH6: Only carry out switching operation SH6 if the state of the circuit between the phase conductor to be switched on according to switching operation SH6 and the neutral conductor is “normal”.

Additional switching-off actions are defined as:

• Switching operation SH7: a. if the power semiconductor is switched on in only one phase (switching state SZ1), and b. the magnitude of the phase voltage between a phase conductor with the power semiconductor switched off and the neutral conductor is less than a certain value uMAXLN(off), and c. the state of the (linked) circuit between this phase conductor and the phase conductor already switched on is “overcurrent”, switch off the power semiconductor of the switched-on phase, whereby the value uMAXLN(off) is greater than uMAXLN(on).

(This is followed by automatic switching on of the phase conductor mentioned under b) by switching operation SH1)

• Switching operation SH8: a. if power semiconductors are switched on in only two phases (switching state SZ2), and b. the magnitude of the phase voltage between the phase conductor with the power semiconductor switched off and the neutral conductor is less than a certain value uMAXLN(off), and c. the state of the interlinked circuit between this phase conductor and the phase conductor already switched on is “overcurrent”, switch off the power semiconductor of this switched-on phase, where the value uMAXLN(off) is greater than uMAXLN(on).

• Switching operation SH9: a. if power semiconductors in two phases are switched on (switching state SZ2), and b. the amount of the interlinked voltage between the phase conductor LX with the power semiconductor switched off and a second phase conductor LY with the power semiconductor switched on is less than a certain value uMAXLL(off), and c. the state of the interlinked circuit between the phase conductor with the power semiconductor switched off and the third phase conductor LZ with the power semiconductor switched on is “overcurrent”, switch off the power semiconductor in the phase conductor LZ, where the value uMAXLL(off) is greater than uMAXLL(on).

The useful range of values for the values uMAXLN(off) and uMAXLL(off) is similar to that for the values uMAXLN(on) and uMAXLL(on). Setting uMAXLN(off) greater than uMAXLN(on) and uMAXLL(off) greater than uMAXLL(on) ensures that Phase is switched off and then another phase is switched on after a certain time interval.

The cause of the overcurrent can be determined using the current values before the overcurrent event. For this purpose, it is useful to define virtual circuits (SC) as follows: o SKL1 N between the phase conductor L1 and the neutral conductor N o SKL2N between the phase conductor L2 and the neutral conductor N o SKL3N between the phase conductor L3 and the neutral conductor N o SKL1 L2 between the phase conductor L1 and the phase conductor L2 o SKL2L3 between the phase conductor L2 and the phase conductor L3 o SKL3L1 between the phase conductor L3 and the phase conductor L1

In addition, two circuit states are defined: each of the circuits SKL1 N, SKL2N, SKL3N, SKL1 L2, SKL2L3, SKL3L1 has the state o "normal" if the probability of an overcurrent event is estimated to be low when the respective circuit is connected to the grid. o "overcurrent" if the probability of an overcurrent event is estimated to be high when the respective circuit is connected to the grid.

The following measurements and calculations can then be carried out:

• The phase conductor currents i1, i2, i3 are continuously measured or sampled.

• In addition to measuring the phase currents, their values from one or more points in the past are stored in such a way that after an overcurrent event in one phase conductor, current values from all three phases are available for analysis, which were measured immediately before this overcurrent event, i.e. before the current in the affected phase conductor was interrupted by switching off the power semiconductor.

• Calculation rule for calculating the virtual chained currents through the circuits SKL1 L2, SKL2L3, SKL3L1: o The virtual chained current iSKL1 L2 in the circuit SKL1 L2: iSKL1 L2 = I *(i1 - i2) o The virtual chained current iSKL2L3 in the circuit SKL2L3: iSKL2L3 = %*(i2 - i3) o The virtual chained current iSKL3L1 in the circuit SKL3L1: iSKL3L1 = %*(i3 - i1)

• Calculation rule for the continuous calculation of the virtual neutral conductor current iN: o iN = k*(i1 + i2 + i3) with k as weighting factor, k = 1...2. Setting the circuit states to “normal”:

• For each of the circuits SKL1 N, SKL2N, SKL3N, the circuit state is set to “normal” as soon as the power semiconductor is switched on in the respective phase and no overcurrent event occurs.

• For each of the circuits SKL1L2, SKL2L3, SKL3L1, the circuit state is set to “normal” as soon as the power semiconductors are switched on in the two phases connected to the circuit and no overcurrent event occurs in any of the connected phases.

Setting the circuit states to “overcurrent”:

• In the event of an overcurrent event in a phase conductor LX, the state of the circuit LXN between the phase conductor LX and the neutral conductor N is set to “overcurrent” if immediately before this overcurrent event the magnitude of the virtual neutral conductor current iN was greater than o the magnitude of the virtual interlinked current iLXLY in the circuit between this phase conductor LX and a second phase conductor LY, o the magnitude of the virtual interlinked current iLZLX in the circuit between this phase conductor LX and the third phase conductor LZ.

• In the event of an overcurrent event in a phase conductor LX, the state of the circuit between the phase conductor LX and a second phase conductor LY is set to “overcurrent” if immediately before this overcurrent event the magnitude of the virtual interlinked current iLXLY in the circuit SKLXLY was greater than o the magnitude of the virtual neutral conductor current iN and o the magnitude of the virtual interlinked current iLZLX between this phase conductor LX and the third phase conductor LZ

• The indices X, Y, Z stand for 1, 2, 3 or 2, 3, 1 or 3, 2, 1.

The following should be noted:

• The virtual circuits SKL1N, SKL2N, SKL3N, SKL1L2, SKL2L3, SKL3L1 and the virtual currents through these circuits, as well as the virtual neutral current, serve solely as an abstraction for estimating current flows in the load and are always calculated, regardless of the actual connected loads. The goal of this abstraction is to consider all potential current flows in load monitoring.

• An accurate assessment of the current flows in the load, ie an accurate determination of the currents in the virtual circuits, is not possible and is not achieved for all load configurations. • Shortly before an overcurrent event, the inrush or fault current is many times greater than the normal operating current. The current pattern in the four-wire system is dominated by the inrush or fault current, with the inrush or fault current almost always flowing through only one of the six circuits.

• This makes this approach valuable for the purpose of determining the cause of overcurrent, although it is only partially suitable for analysis in a normal operating condition.

With these considerations, the cause of the overcurrent can be determined using the impedance values before the overcurrent event. The following continuous measurements and calculations are performed:

• The phase currents i1, i2, i3 are continuously measured or sampled by the SCCB.

• The phase voltages u1, u2, u3 are continuously measured or sampled by the SCCB.

• In addition to measuring the phase currents and phase voltages, their values from one or more points in the past are stored in such a way that after an overcurrent event in one phase conductor, current and voltage values from all three phases are available for analysis, which were measured immediately before this overcurrent event, i.e. before the current in the affected phase conductor was interrupted by switching off the semiconductor.

• Calculation rule for the continuous calculation of the impedances zL1N, zL2N, zL3N of the virtual circuits SKL1 N, SKL2N, SKL3N: o Impedance of the circuit L1 N: zL1 N = uL1 N/ i1 o Impedance of the circuit L2N: zL2N = uL2N/ i2 o Impedance of the circuit L3N: zL3N = uL3N/ i3

• Calculation rule for the continuous calculation of the impedances zL1L2, zL2L3, zL3L1 of the virtual circuits SKL1 L2, SKL2L3, SKL3L1: o Impedance of the circuit SKL1 L2: zL1L2 = (uL1 N - uL2N)/(1/2*(i1 - i2)) o Impedance of the circuit SKL2L3: zL2L3 = (uL2N - uL3N)/(1/2*(i2 - i3)) o Impedance of the circuit SKL3L1: zL3L1 = = (uL3N - uL1 N)/(1/2*(i3 - i1))

Setting the circuit states to “normal”:

• For each of the circuits SKL1 N, SKL2N, SKL3N, the circuit state is set to “normal” as soon as the semiconductor is conductive in the respective phase and no overcurrent event occurs. • For each of the circuits SKL1 L2, SKL2L3, SKL3L1, the circuit state is set to “normal” as soon as the semiconductors in the two phases connected to the circuit are conductive and no overcurrent event occurs in any of the connected phases.

Setting the circuit states to “overcurrent”:

• In the event of an overcurrent event in a phase conductor LX, the state of the circuit LXN between this phase conductor LX and the neutral conductor N is set to “overcurrent” if the following conditions were met simultaneously immediately before this overcurrent event: o the power semiconductor of phase LX was switched on o the impedance zLXN was positive o if power semiconductors of a second phase LY were switched on: the impedance zLXLY was positive and greater than impedance zLXN o if power semiconductors of a third phase LZ were switched on: the impedance zLZLX was positive and greater than impedance zLXN

• In the event of an overcurrent event in a phase conductor LX, the state of the circuit between this phase conductor LX and a second phase conductor LY is set to “overcurrent” if the following conditions were met simultaneously immediately before this overcurrent event: o the power semiconductors of phases LX and LY were switched on o the impedance zLXLY was positive o the impedance zLXN was positive and greater than the impedance zLXLY o if power semiconductors of the third phase LZ were switched on: the impedance zLZLX was positive and greater than the impedance zLXLY

• The indices X, Y, Z stand for 1, 2, 3 or 2, 3, 1 or 3, 2, 1.

The following should be noted:

• The virtual circuits SKL1 N, SKL2N, SKL3N, SKL1 L2, SKL2L3, SKL3L1 and the circuit states are defined as above.

• Analogous to the consideration of the virtual currents, the calculated virtual impedances zL1 N, zL2N, zL3N, zL1 L2, zL2L3, zL3L1 are an abstraction for the assessment of the load state and are particularly valuable shortly before an overcurrent event.

• Impedance analysis can be advantageous over current analysis when high currents occur simultaneously between a phase conductor and the neutral conductor and between the same phase conductor and a second phase conductor. • Determining the probable cause of an overcurrent based on impedances increases the probability of correctly determining the overcurrent. This allows for better switching decisions.

The following describes the conditions for permanent shutdown. To complete the operating procedure, conditions for terminating the reclosing attempts must be determined. Three options for such conditions are described below, each of which may be advantageous from different perspectives:

Conditions for switching off based on the maximum number of switching attempts for a circuit

Conditions for dimensioning the energy absorbers

Conditions for compliance with a standard or regulation on the duration of the ultra-short voltage interruption. i) Permanent shutdown after a number of switch-on attempts with overcurrent events

1. For each of the circuits SKL1 N, SKL2N, SKL3N, SKL1 L2, SKL2L3, SKL3L1, an overcurrent event counter is initialized to zero.

2. Initially, the SCCB is in the "stationary" state. This refers to a. the state in which all three power semiconductors are switched on and the loads are continuously supplied without any overcurrent events occurring. b. the state in which all power semiconductors are switched off before an external switch-on request to the SCCB is received.

3. In the event of an overcurrent event, the most probable overcurrent cause is assigned to one of the circuits SKL1N, SKL2N, SKL3N, SKL1 L2, SKL2L3, SKL3L1 according to the procedure described above. The overcurrent event counter of the circuit to which the

Overcurrent cause is incremented.

4. If one of the overcurrent event counters reaches a certain limit, the SCCB is permanently deactivated. It is advantageous to: a. Set different limit values for the group of counters for circuits SKL1 N, SKL2N, and SKL3N than for the group of counters for circuits SKL1 L2, SKL2L3, and SKL3L1. b. Set a single limit value for the sum of the counters SKL1L2, SKL2L3, and SKL3L1.

5. After all three power semiconductors have been switched on, the counting of the time TSTAT begins. 6. After the time TSTAT exceeds a certain value TSTAT.MAX, the SCCB returns to the "stationary" state. The overcurrent event counters are reset. ii) Permanent shutdown after a certain number of overcurrent events in a phase

1. For each phase power semiconductor, an overcurrent event counter is initialized to zero.

2. Initially, the SCCB is in the "stationary" state. This refers to a. the state in which all three power semiconductors are switched on and the loads are continuously supplied without any overcurrent events occurring. b. the state in which all power semiconductors are switched off and a switch-on request to the SCCB comes from outside.

3. After an overcurrent event occurs in a phase, the overcurrent counter of that phase is incremented.

4. After all three power semiconductors have been switched on, the counting of the time TSTAT begins.

5. After the time TSTAT exceeds a certain value TSTAT.MAX, the SCCB returns to the "stationary" state. The overcurrent event counters are reset. iii) Permanent shutdown after a time interval

1. Initially, the SCCB is in the "stationary" state. This refers to a. the state in which all three power semiconductors are switched on and the loads are continuously supplied without any overcurrent events occurring. b. the state in which all power semiconductors are switched off before an external switch-on request to the SCCB is received.

2. After an overcurrent event occurs, the procedure is executed for a certain time TMAX.

If power semiconductors of all three phases are not switched on after TMAX has elapsed, the SCCB is permanently switched off and is in the “Fault” state.

3. As long as overcurrent events occur during the time TMAX, the SCCB is in the “Transient” state.

4. As soon as all three power semiconductors are switched on, the counting of the time TSTAT begins.

5. If an overcurrent event occurs within the time TMAX, the counter for TSTAT is reset to zero and incremented again the next time the state with all three switches switched on is reached. 6. If an overcurrent event occurs outside the TMAX time and within the TSTAT time, the SCCB is permanently switched off.

7. After the time TSTAT exceeds a certain value TSTAT.MAX, the SCCB returns to the "stationary" state.

The inventive approach is explained in detail below using operating scenarios. All figures described below are created in a simulation and serve to illustrate the behavior of the SCCB in the described operating procedures. Noise and peaks in the voltage time profiles are caused by the simulation parameters.

Scenarios 1 to 14 demonstrate the current and voltage waveforms of an SCCB with three single-phase resistive loads. Short circuits are connected in parallel if necessary.

The first scenarios concern basic switching procedures.

The first scenario relates to switching on without overcurrent events, beginning with switching operation SH1, and is shown in Fig. 4. It illustrates how, with all three switches switched off (switching state SZO), the power semiconductor of phase 3 is first switched on by switching operation SH1. Afterward, current flows only between the switched-on phase 3 and the neutral conductor. In this case, the current in the virtual circuit SKL3N corresponds exactly to the current in phase conductor 3.

As long as no overcurrent event occurs, the device is in switching state SZ1, and the next switching action is the switching on of the power semiconductor of a second phase if the magnitude of the phase-to-phase voltage between the switched-on phase and the other phase is small enough. In scenario 1, the switching on of the power semiconductor of phase 2 occurs with the switching action SH4. After switching on phase 2, currents flow in phases 2 and 3. Currents can flow in the virtual circuits SKL2N, SKL3N, and SKL2L3, if the circuits exist.

Subsequently, switching operation SH6 is executed from switching state SZ2, in which the power semiconductor of the still switched-off phase 1 is switched on at a zero crossing of the voltage between phase conductors 1 and 2. The device is in switching state SZ3, and all possible loads are continuously supplied. The second scenario concerns a switch-on without overcurrent events starting with switching operation SH2.

Depending on the time at which the switch-on request occurs in the switching state SZO, the switching operation SH1 or SH2 can be executed first in the switching state SZO. Fig. 5 shows an example in which the switching operation SH2 is executed first. If no overcurrent event occurs, the device is in the same state after this switching operation as after the switching operation SH4 in Scenario 1 described above. The subsequent sequence is similar to Scenario 1.

The third scenario concerns a switch-on with a short circuit in phase 3.

If there is a short circuit in the SKL3N circuit, an overcurrent event will occur after the power semiconductor in phase 3 is switched on. Scenario 3 in Fig. 6 shows such a case. Similar to Scenario 1, the power semiconductor in phase 3 is switched on (switching operation SH1), but shortly afterwards the excessive current is interrupted with the subsequent switching operation SHO. After the switching operation SHO, the switching operation SH2 is carried out near the zero crossing of the voltage between phase conductors 2 and 3. Since the power semiconductor in phase 3 is not switched on at the zero crossing of the voltage uL3N during the switching operation SH2, the steepness of the short-circuit current is higher than in the earlier switch-on attempt and an overcurrent event and a switching operation SHO occur more quickly.

The fourth scenario concerns switching on with a short circuit in phase 2.

Fig. 4 shows how, in the event of a short circuit in the SKL2N circuit after the SH4 switching operation, overcurrent events occur in phase conductor 2. From the SH1 switching operation to the SH4 switching operation, the current and voltage curves are similar to those in scenario 3. After the SH4 switching operation, the SH4 switching operation interrupts the overcurrent in phase 2, and the device is in the SZ1 switching state with phase 3 switched on. From this state, the SH3 switching operation is performed to switch on phase 2 at the zero crossing of the voltage uL3N.

The fifth scenario concerns switching on with a short circuit between phases 2 and 3. A short circuit in the circuit between two phases causes power semiconductors in one or both affected phases to be switched off. Scenario 5 in Fig. 8 shows such a case (short circuit between phases 2 and 3, i.e. in the SKL2L3 circuit). First, the power semiconductor in phase 3 is switched on by switching operation SH1. After switching operation SH4, an overcurrent occurs in phase 2 and the power semiconductor in phase 2 is switched off by switching operation SHO. Then, when the phase voltage uL2N crosses zero, the power semiconductor in phase 2 is switched on again by switching operation SH3. After this switching operation, the current in phases 2 and 3 rises with a faster gradient because the voltage uL2L3 between the two phase conductors is significantly higher than after switching operation SH4. However, during this overcurrent event, the power semiconductor in phase 3 is switched off because, due to the different voltage conditions, the current in this phase reaches the current limit more quickly. Thus, between switching operations SH1 and SH4, a current flows through phase conductor L1; from switching operation SH3 until switching operation SH4 at the earliest, a current flows through phase conductor L2. If single-phase rectifiers are connected to phases 2 and 3, the intermediate circuit capacitances of the two rectifiers can be partially charged during these time intervals. And if the overcurrent between phases 2 and 3 is not caused by a short circuit but by a three-phase rectifier, then the intermediate circuit capacitance of this rectifier will also be partially charged.

The sixth scenario concerns the execution of switching operation SH 5.

Scenario 6 is an example of the execution of the switching operation SH5 (Fig. 9). During operation in the switching state SZ3 (the power semiconductors of all three phases are switched on), an overcurrent occurs in phase 1, e.g. due to the switching on of an inrush or short-circuit load with an overcurrent between the phase conductor L1 and the neutral conductor N. The time of switching on is before a zero crossing of the phase voltage uL1N and after a zero crossing of one of the line-to-line voltages uL1 L2 or uL3L1. After the short-circuit is switched on, the power semiconductor of phase 1 is switched off due to the overcurrent with the switching operation SHO. It is then switched on at the zero crossing of the phase voltage uL1N by the switching operation SH5 and shortly afterwards switched off again by the switching operation SHO. This is followed by switching on at the zero crossing of the voltage uL3L1 and repeated switching off by the switching operation SHO. Scenarios 1 to 6 are briefly summarized below. The scenarios described above demonstrate how the basic switching procedure systematically attempts to utilize every zero voltage crossing for the individual switching on and off of the power semiconductors. If inrush loads are connected to the phases instead of the short circuits shown, this switching tactic transfers energy to these loads in the "gentlest" way possible. The procedure is optimal for switching single- or three-phase loads on and off as quickly as possible.

In some load cases, switching on a power semiconductor at the zero crossing of a phase voltage is not always optimal for a linked inrush or short-circuit load, and vice versa. The most complex load case occurs when single-phase and two- or three-phase loads are connected simultaneously. For this reason, the execution of the switching operations can be improved to avoid some switch-on commands if, based on previous tests, an overcurrent event is expected to occur subsequently. Additionally, in some cases, briefly switching off a power semiconductor can be advantageous.

The following scenarios illustrate additional conditions for switching on when single-phase overcurrents are detected.

The seventh scenario concerns switching on with two single-phase short circuits in phases 3 and 1 using the basic switching procedure.

The switching operation SH2 in Scenario 3 (Fig. 6) can be omitted if a previous switch-on attempt detected that switching on this phase would trigger an overcurrent between this phase and the neutral conductor. To demonstrate the advantages of such a tactic, Scenarios 7 and 8 are shown in Fig. 10 and Fig. 11.

In scenario 7 (see Fig. 10), a low-resistance short circuit exists between phase conductors 1 and 3 and the neutral conductor. The short circuit in phase 1 is simulated only to bring the SCCB into a state where the switching operations SH3 and SH4 in phase 3 can be demonstrated immediately after SH1 and SH2. The switching operations in phase 1 are not considered and are therefore shown in gray.

After switching on phase 3 by SH1 at the time shortly before 0.148 [s], an overcurrent shutdown with SHO follows. This is followed by switching on phases 2 and 3 by the Switching operation SH2. A short time later, phase 3 is again switched off due to an overcurrent by SHO. Switching phase 3 on again by SH2 at an even higher phase voltage than during the first switch-on attempt (SH1) was not advisable. Omitting SH2 for phase 3 can prevent unnecessary stress on the energy absorber when switching off the overcurrent. Omitting switching operation SH4 after 0.158 [s] is advantageous for the same reasons. The process can be improved by additional conditions.

The eighth scenario shows switching on with two single-phase short circuits in phases 3 and 1 using the basic switching procedure and additional conditions ZB1 and ZB2.

Scenario 8 of Fig. 11 shows how, by introducing the additional conditions ZB1 and ZB2 for the switching operations SH2 and SH4, the unnecessary switching operations are avoided, in contrast to Scenario 7. Thus, with the power semiconductor in phase 3, only two instead of four overcurrent events had to be switched off in the shown period between 0.148 and 0.158 [s] (half the grid period) compared to Scenario 7. This is a clear advantage for the dimensioning of the energy absorber of an SCCB.

The ninth scenario concerns switching on with a single-phase short-circuit in phase 3 using the basic switching procedure and additional conditions SZ1 and SZ2.

Scenarios 9 and 10 in Fig. 12 and Fig. 13 serve to demonstrate the effect of the additional condition ZB3.

In scenario 9 of Fig. 12, only phase conductor 3 is short-circuited to the neutral conductor. Thus, after time 0.156 [s], power semiconductors in two phases are switched on, and SH6 occurs in phase 3 (in contrast to SH4 in scenario 7). This is followed by an overcurrent shutdown with SHO. The switching operation SH6 is superfluous for the same reasons as the switching operations SH2 and SH4 in scenario 7.

Scenario ten concerns switching on with a single-phase short circuit in phase 3 with the basic switching procedure and additional conditions SZ1, SZ2 and SZ3.

Scenario 10 of Fig. 13 shows the omission of the switching operation SH6 when introducing the additional condition ZB3 compared to Scenario 9. In summary, the following can be said about scenarios 7 to 10. With the avoidable switch-on operations SH2, SH4, and SH6, the circuits SKL1N, SKL2N, and SKL3N are switched on at times when the voltage across the respective circuit is equal to half the phase voltage amplitude. The benefit of switching on these circuits at such times is very minimal. If the overcurrents are caused by a capacitance, these times are not optimal for switching on and charging this capacitance. Instead, the times near the phase voltage zero crossings should be used (switching operations SH1, SH3, SH5). These times are also more suitable as switch-on moments for analyzing and distinguishing a short circuit from an inrush load.

On the other hand, the additional conditions ZB1, ZB2, and ZB3 prevent up to two shutdowns per half grid period in each phase. This reduces the amount of energy absorbed by the energy absorber by fifty percent.

The following scenarios concern proactive shutdown when two-phase overcurrents are detected.

The eleventh scenario concerns switching on with a two-phase short circuit between phases 2 and 3 with the basic switching procedure and additional conditions SZ1, SZ2 and SZ3.

If there is an overcurrent load between two phases (circuits SKL1L2, SKL2L3, SKL3L1), overcurrent events can occur after switching operations SH3 and SH5. Scenario 11 of Fig. 14 shows an example. First, switching operations SH1 and SH2 are performed; shortly after SH2, an overcurrent occurs between phases 2 and 3, and phase 2 is switched off via SHO. Then, phase 2 is switched on via SH3, which again causes an overcurrent between phases 2 and 3. Using the SH2 switching operation and the subsequent SHO switching operation, it can be determined that the circuit between phases 2 and 3 has a high probability of generating overcurrent events. With this information, it is already known before the SH3 switching operation that it would lead to an overcurrent, because the voltage between phases 2 and 3 is significantly higher shortly before the SH3 switching operation than during the earlier attempt to switch these two phases on simultaneously (SH2).

From the point in time immediately before the switching operation SH3, four cases are basically possible: 1. Switching on the power semiconductor in phase 2 via SH3 and then switching off phase 2 triggered by the overcurrent, as shown in Fig. 14. In this case, phase 3 continues to be energized. At time just before 0.155 [s], phase 1 is switched on via SH3 and can also be energized. Then, at time 0.16 [s], SH6 is executed and the circuit between phases 2 and 3 is switched on again.

2. Switching on the power semiconductor in phase 2 by SH3 and then switching off phase 3 triggered by the overcurrent (not shown in Fig. 14). In this case, single-phase loads in circuits SKL2N, SKL1L2, and SKL3N could switch on "softly" one after the other as quickly as possible at the relevant voltage zero crossings. (Whether phase 2 or 3 switches off during the two-phase overcurrent depends on the load conditions and the current measurement tolerances and offsets.)

3. Switching on the power semiconductor in phase 2 by SH3 and subsequently switching off both phases 2 and 3 due to the overcurrent (not shown in Fig. 14). In this case, depending on the computational implementation, the algorithm would either a. immediately switch on the power semiconductor in phase 2, in which case the sequence according to case 2 applies, or b. wait until the phase voltage uL1 N crosses zero to switch on phase 1. This means that for approximately one-sixth of the grid period, no loads are supplied and no attempts are made to switch on new circuits.

4. Proactively switch off phase 3 and then switch on phase 2 via switching operation SH1 (not shown in Fig. 14). In this case, the same loads are supplied as in case 2. The difference is that during proactive switch-off, the power semiconductor of phase 3 does not switch off an overcurrent, but rather a smaller operating current.

In cases 2, 3.a and 4, the energy can be “portioned” among the circuits: all available voltage zero crossings are optimally used for switching on the circuits.

In the other cases, the loads are supplied with energy asymmetrically or even all three phases are switched off for about 3.33 [ms].

Symmetrical precharging of the loads in circuits SKL1 N, SKL2 N, and SKL3 N (phase-neutral) is a prerequisite for the smooth start-up of the other circuits (phase-phase) and must therefore be given higher priority. From this perspective, case 4 is the most attractive. Furthermore, it is the most favorable for the load on the energy absorbers. The twelfth scenario concerns a demonstration of the switching operation SH7.

Scenario 12 of Fig. 15 shows the proactive switching off of phase 3 with the switching action SH7, as described in case 4 above.

The thirteenth scenario concerns a demonstration of the switching operation SH8.

In Scenario 12, the switching operation SH7 is performed in switching state SZ1. It is also advantageous to proactively switch off a phase in switching state SZ2 when two-phase overcurrent loads are detected. Scenario 13 of Fig. 16 shows such a case. The switching operations before time 0.16 [s] are the same as in Scenario 12. The difference is that after 0.16 [s], the switching operation SH8 is performed instead of SH5. This switches off a smaller current than the overcurrent after SH5 in Scenario 12. Furthermore, in Scenario 13, the later switching operations SH6 and SHO do not occur, unlike in Scenario 12.

In contrast to scenario 12, executing SH8 results in phase 2 being energized in the time period after 0.16 [s].

The switching operations SH7 and SH8 ensure that the circuits are alternately energized in switching states SZ1 and SZ2.

The fourteenth scenario concerns a demonstration of the switching operation SH9.

In scenarios 12 and 13, after the switching operation SH6, an overcurrent event occurs shortly after the time 0.155 [s], after which phase 2 is switched off. With the switching operation SH6, phase 2 is switched on, whereby

• Phase 3 is switched on

• it is known that an overcurrent occurred between phases 2 and 3 during a previous switch-on attempt near the relevant zero crossing

• The voltage between phases 2 and 3 shortly before the switching operation SH6 mentioned is approximately 3/2 of the phase voltage amplitude.

It is therefore to be expected that after switching on phase 2 by SH6, an overcurrent will occur between phases 2 and 3. From the time before SH6, the following cases are possible: 1. Execution of SH6 and deactivation of phase 2 due to overcurrent, as shown in scenarios 12 and 13. This allows the circuits SKL1N, SKL3N, and SKL3L1 to be energized after the SHO switching operation in the time interval 0.155-0.16 [s]. The zero crossings of the voltages uL3L1 and uL3N can be optimally utilized for reclosing in the event of further overcurrent events.

2. Execution of SH6 and deactivation of phase 3 due to overcurrent. Thus, in the time period following the switching operation SHO, the SKL2N circuit is energized for a period of 0.155-0.16 [s].

3. Proactively switching off phase 2 via switching operation SH9, as shown in Fig. 17. In this case, the current waveforms after the subsequent switching on of phase 3 are analogous to case 1. The difference is that a smaller current is switched off from phase 2. Cases 1 and 3 are advantageous for evenly flowing current to the potential inrush loads. Only case 3 can be deliberately induced. If phase 2 is not proactively switched off, it is unknown whether case 1 or 2 will occur afterwards. If the previous overcurrent between phases 2 and 3 was an inrush current from a three-phase rectifier, it will most likely no longer have a high amplitude when phase 3 is switched on by SH6 (scenario 13), since the intermediate circuit was pre-charged by the previously switched-on phases 1 and 2.

Carrying out the switching operation SH6 as in scenarios 12 and 13 is advantageous for switching on all phases as quickly as possible in the case where the overcurrent between phases 2 and 3 is not caused by a two-phase short circuit but by a three-phase rectifier.

Performing the SH9 switching operation instead of SH6 slows down the entire closing process, but offers the possibility of avoiding an overcurrent between phases 2 and 3 if it is caused by a short circuit. Furthermore, in this case, the circuit between phases 3 and 1 can be closed and analyzed.

A control circuit breaker without the SH9 switching action is characterized by a somewhat faster switch-on behavior. A control circuit breaker with SH9 is slower and more "prudent" and can offer advantages from a load detection perspective.

In summary, the following can be stated for scenarios 11 to 14. With the

Switching operations SH7 and SH8 achieve a symmetrical energy distribution between single-phase loads in transient conditions. This is an advantage and prerequisite for the connection of two- and three-phase loads. The SH9 switching operation is optional and is only relevant under special circumstances, for example, when a longer on- or off-load is required.

Re-switching is permitted and detection of the load generating the overcurrent is desired.

In the following, options for designing the control system are discussed.

The basic switching procedure consists of switching operations SHO to SH6 and aims to utilize the zero crossing of each of the six voltages for switching the respective circuit on or off. The minimum equipment of a control device must include the basic switching procedure and a criterion for permanent switching off.

The additional conditions ZB1 to ZB3 and switching operations SH7 to SH9 are advantageous additions to the basic switching procedure. Their implementation requires monitoring of the circuit states.

Two methods for monitoring the circuit states and determining the

The causes of overcurrent are described above. It cannot be ruled out that variations of these procedures or other procedures could be implemented for this purpose. Regardless of the monitoring implementation, it is important that after an overcurrent event, the cause of the overcurrent is assigned to one of the six virtual circuits. After this, it is expected that the next time one or two switches are switched on, an overcurrent event will occur in a specific circuit with a very high probability.

The additional conditions ZB1 to ZB3 improve the basic switching procedure by omitting switching-on operations that would most likely only lead to an overcurrent shutdown of the power semiconductor to be switched on.

The switching operations SH7 and SH8 proactively ensure that the energy is distributed as evenly as possible among the loads during the switch-on and switch-back processes. This may result in the fully switched-on state SZ3 being reached more quickly.

The switching operation SH9 is optional. Individually and together, the additional conditions ZB1 to ZB3 switching operations SH7 to SH9 offer an opportunity to positively influence the dimensioning of the energy absorbers.

The benefits of these additions are only visible in certain operating conditions. In simple operating cases, they are hardly noticeable, e.g., with simple resistive loads without overcurrent.

Examples of a switch-on process with a rectifier load are described below.

The fifteenth scenario in Fig. 18 shows the switching on of an SCCB using the basic switching method. The loads are three single-phase rectifiers with 400 pF DC link capacitance each and one three-phase rectifier with 200 pF DC link capacitance.

The sixteenth scenario in Fig. 19 demonstrates the switching on of an SCCB with the basic switching procedure supplemented by the additional conditions ZB1 to ZB3 and switching operations SH7 and SH8 with the same loads. The transient process takes a few [ms] longer in scenario 16 than in scenario 15. In both scenarios, all three phases are permanently switched on after less than one grid period from the first switching operation.

Fig. 20 shows a first basic procedure that provides for the permanent switching off of a circuit breaker according to the invention if stable operation is not achieved. This first procedure is based on counting switching events using a counter. The method is started in step S11, e.g., as part of a switch-on routine of a circuit breaker according to the invention. The counter is set to zero (step S12) and the registration of overcurrent events is started (step S13). During operation, stable operation may occur, which triggers a reset of the counter (step S14). The procedure for a reset is shown in more detail in Fig. 22. In the event of an overcurrent event S15, the counter is incremented. If a threshold value for the counter reading is reached, this leads to the permanent switching off (step S18). Otherwise, the method continues. The S14 reset ensures that a permanent shutdown does not simply occur due to the accumulation of overcurrent events, but only in the event of unstable operation.

Fig. 21 shows a second basic procedure which provides for a permanent switching off of a circuit breaker according to the invention if stable operation is not achieved. This method operates with a timer. During the initialization of the switch (step S21), the detection of overcurrent events is provided (step S22). If an overcurrent event occurs (step S23), a timer is started (step S24). While this timer is running, a reset can occur (step S25). The reset is explained in more detail below using Fig. 22. The reset terminates the timer (step S26). A new timer can be started if another overcurrent event occurs. If the timer reaches a limit value Tmax (step S27), this leads to a permanent shutdown (step S28).

The methods shown in Fig. 20 and Fig. 21 can also be used in parallel or in combination.

Fig. 22 shows a reset method that can be used for both counter-based (Fig. 20) and timer-based (Fig. 21) permanent shutdown. If an overcurrent event S32 and the interruption of a phase line occur after the switch has been started (step S31), the switch is restarted using one of the switching operations described above (step S33). When all phases have been switched on by switching operations (query S34), a timer Tstat is started (step S35). This timer is cleared (step S37) if one of the three phases is interrupted again (due to an overcurrent event) (step S36). A corresponding timer would then be initiated again when the three phases are connected again. If the timer reaches a limit value Tstat, max (step S38) without any of the phases being interrupted, the criterion for stable operation is met and a reset is triggered in the process of Fig. 20 or Fig. 21 (step S39).

A novel approach to operating a 3+N SCCB with control of inrush currents by controlling the power semiconductors is described above. This approach is less complex and more efficient than conventional approaches known from other switch types. The embodiments are merely illustrative. Numerous modifications that fall within the scope of the claims will be readily apparent to those skilled in the art.

Claims

Patent claims 1. Circuit breaker for protecting a power supply formed by three phase conductors (L1, L2, L3) and one neutral conductor (N), the circuit breaker - for determining current values of the three phase conductors (L1, L2, L3), - for determining the values of the voltages between the phase conductors (L1, L2, L3) and the neutral conductor (N), - for the evaluation of certain current and voltage values with a view to fulfilling conditions for switching operations involving the interruption of the phase conductors (L1, L2, L3) and the restoration of interrupted connections, - to output control signals in accordance with switching operations to be carried out according to the evaluation results, and - is designed to interrupt the phase conductors (L1, L2, L3) and restore interrupted connections by means of the control signals, whereby - the circuit breaker is designed to use a criterion for a possible overcurrent when carrying out at least one switching operation to be carried out, -- to determine from the current values determined whether this criterion is met, and -- to carry out the switching operation in accordance with this criterion. 2. Circuit breaker according to claim 1, characterized in that - a switching operation comprises the restoration of at least one interrupted connection of a phase conductor, and - the circuit breaker is designed not to carry out the switching operation if the criterion is met. 3. Circuit breaker according to claim 2, characterized in that - a switching operation involves the restoration of an interrupted connection of a phase conductor, - a virtual circuit is defined for the phase conductor and the neutral conductor (N), and - the circuit breaker is designed not to carry out the switching operation if the criterion for a possible overcurrent on this virtual circuit is met. 4. Circuit breaker according to claim 3, characterized in that - a switching operation involves restoring the interrupted connections of two phase conductors, - two virtual circuits are defined for each of the phase conductors and the neutral conductor, and - the circuit breaker is designed not to carry out the switching operation if the criterion for a possible overcurrent is met for one of the virtual circuits. 5. Circuit breaker according to one of claims 1 to 4, characterized in that - a switching operation involves the interruption of at least one phase conductor, and - the circuit breaker is designed to carry out the switching operation when the criterion is met. 6. Circuit breaker according to claim 5, characterized in that - a virtual circuit with another phase conductor is defined for the phase conductor and the criterion for possible overcurrent is met for this virtual circuit, - the other phase conductor is disconnected but meets a sufficient condition for reconnection. 7. Circuit breaker according to one of the preceding claims, characterized in that - a switching operation to be carried out in accordance with a criterion for possible overcurrent is assigned to a virtual circuit, and - the circuit breaker is designed to -- to calculate the current values characterising the virtual circuit from the measured current values, and -- to consider the criterion for a possible overcurrent as fulfilled if (i) a phase conductor of the virtual circuit has been interrupted due to an overcurrent event, and (ii) the current values characterising the virtual circuit immediately before the interruption of the phase conductor fulfilled at least one condition for overcurrent. 8. Circuit breaker according to claim 7, characterized in that - it is a virtual circuit with a phase conductor and the neutral conductor (N), - at least one condition is that the following is cumulatively fulfilled: the absolute value of the weighted sum of the current values of the three phases is greater than the absolute values of the weighted differences between the current values of the phase conductor and the current values of the other two phase conductors. 9. Circuit breaker according to claim 7, characterized in that - it is a virtual circuit with a first phase conductor and a second phase conductor, the first phase conductor being interrupted due to an overcurrent event, - at least one condition is that the following is cumulatively fulfilled: the amount of the weighted difference between the current values of the phase conductors of the virtual circuit is greater than the amount of the weighted sum of the current values of the three phase conductors and greater than the amount of the weighted difference between the current values of the first phase conductor and the third phase conductor not belonging to the virtual circuit. 10. Circuit breaker according to one of the preceding claims, characterized in that - a switching operation to be carried out in accordance with a criterion for possible overcurrent is assigned to a virtual circuit, and - the circuit breaker is designed to set the criterion for a possible overcurrent for this virtual circuit as not fulfilled if all phase conductors of the virtual circuit are conductive without overcurrent. 11. Circuit breaker according to one of the preceding claims, characterized in that - various switching operations are defined for the circuit breaker, which include the re-closing of at least one phase conductor, and - the circuit breaker is designed to carry out only a subset of these switching operations in accordance with a criterion for possible overcurrent. 12. Circuit breaker according to claim 11, characterized in that - switching operations are defined which provide for the restoration of a connection for a phase conductor in accordance with the voltage between the phase conductor and the neutral conductor (N), and - the circuit breaker is designed not to carry out these switching operations in accordance with a criterion for possible overcurrent. 13. A method for protecting a power supply formed by three phase conductors (L1, L2, L3) and one neutral conductor (N), comprising - Determination of current values of the three phase conductors (L1, L2, L3), - Determination of the values of the voltages between the phase conductors (L1, L2, L3) and the neutral conductor (N), - Evaluation of certain current and voltage values with regard to the fulfilment of conditions for switching operations involving the interruption of the phase conductors (L1, L2, L3) and the restoration of interrupted connections, - Output of control signals in accordance with switching operations to be carried out according to the evaluation results, and - interruption of the phase conductors (L1, L2, L3) and restoration of interrupted connections by the control signals, whereby - for at least one switching operation to be carried out -- a criterion for a possible overcurrent is used when carrying out this switching operation, -- it is determined from the current values determined whether this criterion is met, and -- the switching operation is carried out in accordance with this criterion. 14. Method according to claim 13, characterized in that - a switching operation comprises the restoration of at least one interrupted connection of a phase conductor, and - the switching operation is not carried out if the criterion is met. 15. Method according to claim 14, characterized in that - a switching operation involves the restoration of an interrupted connection of a phase conductor, - a virtual circuit is defined for the phase conductor and the neutral conductor (N), and - the switching operation is not carried out if the criterion for a possible overcurrent on this virtual circuit is met. 16. Method according to claim 15, characterized in that - a switching operation involves restoring the interrupted connections of two phase conductors, - two virtual circuits are defined for each of the phase conductors and the neutral conductor, and - the switching operation is not carried out if the criterion for a possible overcurrent is met for one of the virtual circuits. 17. Method according to one of claims 13 to 16, characterized in that - a switching operation involves the interruption of at least one phase conductor, and - the switching operation is not carried out if the criterion is met. 18. Method according to claim 17, characterized in that - a virtual circuit with another phase conductor is defined for the phase conductor and the criterion for possible overcurrent is met for this virtual circuit, - the other phase conductor is disconnected but meets a sufficient condition for reconnection. 19. Method according to one of the preceding claims, characterized in that - a switching operation to be carried out in accordance with a criterion for possible overcurrent is assigned to a virtual circuit, - current values characterising the virtual circuit are calculated from the measured current values, and - the criterion for a possible overcurrent is set as fulfilled if (i) a phase conductor of the virtual circuit has been interrupted due to an overcurrent event, and (ii) the current values characterising the virtual circuit immediately before the interruption of the phase conductor fulfilled at least one condition for overcurrent. 20. Method according to claim 19, characterized in that - it is a virtual circuit with a phase conductor and the neutral conductor (N), - at least one condition is that the following is cumulatively fulfilled: the amount of the weighted sum of the current values of the three phases is greater than the amounts of the weighted differences between the current values of the phase conductor and the current values of the other two phase conductors. 21. Method according to claim 19, characterized in that - it is a virtual circuit with a first phase conductor and a second phase conductor, the first phase conductor being interrupted due to an overcurrent event, - at least one condition is that the following is cumulatively fulfilled: the amount of the weighted difference between the current values of the phase conductors of the virtual circuit is greater than the amount of the weighted sum of the current values of the three phase conductors (neutral conductor current) and greater than the amount of the weighted difference between the current values of the first phase conductor and the third phase conductor not belonging to the virtual circuit. 22. Method according to one of the preceding claims, characterized in that - a switching operation to be carried out in accordance with a criterion for possible overcurrent is assigned to a virtual circuit, and - the criterion for a possible overcurrent for this virtual circuit is set as not fulfilled if all phase conductors of the virtual circuit are conductive without overcurrent. 23. Method according to one of the preceding claims, characterized in that - various switching operations are defined which include the reconnection of at least one phase conductor, and - only a subset of these switching operations is carried out in accordance with a criterion for possible overcurrent. 24. Method according to claim 23, characterized in that - Switching actions are defined that enable the re-establishment of a connection for a Provide phase conductors according to the voltage between the phase conductor and the neutral conductor (N), and - these switching operations are not carried out in accordance with a criterion for possible overcurrent.