WO2023083761A1

WO2023083761A1 - Dc short-circuit current booster

Info

Publication number: WO2023083761A1
Application number: PCT/EP2022/081014
Authority: WO
Inventors: Djamel HADBI; Frédéric REYMOND LARUINA; Philippe EGROT
Original assignee: Electricite De France; Enedis
Priority date: 2021-11-12
Filing date: 2022-11-07
Publication date: 2023-05-19
Also published as: FR3129253A1

Abstract

The invention relates to a device comprising: - an electrical energy storage element defined by an associated DC operating voltage; - a charger configured to charge the storage element when the storage element is discharged and the device is subjected to a DC voltage greater than the DC operating voltage associated with the storage element; and - a diode array defined by an associated threshold voltage value, the diode array being configured to be conductive and thus to discharge the storage element when said storage element is charged and the device is subjected to a DC voltage less than the difference between the DC operating voltage associated with the storage element and the threshold voltage value associated with the diode array.

Description

Title: DC Short Circuit Current Booster

Technical area

This disclosure relates to the field of low voltage public distribution networks, in particular the direct current part (LVDC) and LVDC uses. The present disclosure relates in particular to a device capable of reinforcing the electrical protection of such networks and to a system comprising such a device.

Prior technique

[0002] Interest in direct current networks has been increasing in recent years, particularly in low voltage (LV) networks given the development of native direct current LV receivers, whether in the category of producers (PV), consumers (data center, electric vehicle) or even other uses (stationary storage, micro-network, self-consumption). The main benefits of this technology being better energy efficiency and a certain gain in the Capex of the cables.

[0003] To connect to a direct current network, it is however necessary to use power converters based on semiconductor switches. This raises the issue of protecting this equipment against faults that could occur on the network.

[0004] Figure 1 schematically represents a circuit of an AC/DC converter (1) connected on the one hand to an AC network (2) and on the other hand to a DC outlet (3) which it supplies. The DC output comprises, at least, direct current link capacitors (4) integrated in the AC/DC converter, DC fuses (5) for protection purposes in the event of an electrical fault, a cable (6) which can be modeled by an impedance function of the length of the cable, and a device to be electrically powered (7), which can also be modeled by a resistive load.

[0005] Fuses are a mature and proven technology that is inexpensive and relatively simple. Compared to an AC network whose conventional electrotechnical equipment (transformer, inductors, etc.) can withstand a high current for a duration compatible with the reaction time of fuse or circuit breaker type protection (10 ms), DC networks suffer from several weak points.

[0006] A first weak point is that the fault current does not pass through zero, making it difficult to cut by a circuit breaker.

[0007] A second weak point is the very rapid dynamics of the increase in the fault current (di/dit), which is more severe than in AC faults.

[0008] A third weak point is that the overload capacity and thermal withstand of semiconductor switches is lower than that of electrotechnical equipment (1.2 pu).

Referring to the DC circuit shown in Figure 1 and assuming the appearance of an electrical fault in the cable (6), Figure 2 shows a comparison between, on the one hand, the time of melting of fuses protecting a 40 kW feeder at 750 V DC (i.e. with a nominal current equal to 53 A) and, on the other hand, the destruction time of the semiconductor switches, depending on the distance between the position of the electrical fault and a reference point assimilated to the position of the AC/DC converter.

[0010] More specifically, the curves represented in FIG. 2 correspond to the fusion or destruction times:

- a 63A DC fuse (10),

- an 80A DC fuse (1 1 ),

- an 80A class gR fuse (12),

- an 80A class aR fuse (13),

- a semiconductor switch within a VSC2L type two-level converter (14), and

- a semiconductor switch within a three-level converter of the NPC3L type (15).

[0011] To respond to this problem, two standard strategies can be followed, possibly combined. [0012] One is to develop an electronic protection means that is fast enough to interrupt the fault current before it causes damage to the converters. In this case, the required delay is less than 1 ms.

[0013] One of the known means of protection is the electronic circuit breaker. Such a device is formed by associating in series a sufficiently large number of semiconductor switches sized to block the fault current on a time scale of the order of switching (10 ps). Figure 4 shows an example of such an electronic circuit breaker, comprising two insulated gate bipolar transistors or IGBTs (30, 31), a gate control element (32) and a surge limiter (33). The current path in the event of a fault in an electronic circuit breaker is shown in Figure 4. This technology suffers from several drawbacks starting with the steady state losses in the semiconductors and the high cost. Improvements have been introduced on the basic topology of electronic circuit breakers by associating, in parallel with semiconductor switches (30, 31), a purely electrotechnical part (34), that is to say an ultra-fast switch associated with a lightning arrester. One consequence is to allow nominal operation which bypasses the semiconductor switches, as shown in Figure 5. However, the fault phase response, shown in Figure 6, is then slowed down, reaching around 5 ms (cf. .nplcitl).

[0014] The other standard strategy is to oversize the power converters to give them an overload capacity and a sufficiently high thermal withstand time for the protections conventionally used in AC networks to react, that is to say after a time interval of the order of 10 ms. Work has been carried out to compare the response times of the fastest DC fuses and the destruction time of semiconductor switches subjected to a DC fault, with a 100% oversizing of semiconductor switches compared to the power that the converter is supposed to deliver at rated speed. The results of this work are represented in summary form in Figure 7.

[0015] More specifically, the curves represented in FIG. 7 correspond to the fusion or destruction times: - a DC 63A fuse (40),

- an 80A DC fuse (41 ),

- an 80A class gR fuse (42),

- an 80A class aR fuse (43),

- a semiconductor switch within a VSC2L type two-level converter (44), and

- a semiconductor switch within a three-level converter of the NPC3L type (45).

These results show that a 100% oversizing of the semiconductor switches compared to the power that the converter is supposed to deliver at nominal speed is barely enough to guarantee a sufficiently fast response from the fuses. Moreover, a consequence of such oversizing is a significant increase in the cost of the converters.

[0017] Furthermore, the use of electrostatic energy stored in the capacitors of power converters for the elimination of faults using fuses is well documented in the bibliography. However, with the development of converters switching at high frequencies, a substantial gain in the sizing of the capacitors has been achieved to the detriment of the energy reserve allowing effective and selective protection by fuses (cf. nplcit2). Using the electrostatic energy stored in the capacitors of the power converters to trigger protections of the fuse type is therefore certainly beneficial, but limited in particular because of the oversizing constraints on the capacitors of the power converters (cf. nplcit3), which presents again a significant cost.

Summary

[0018] This disclosure improves the situation.

[0019] A device capable of being inserted into a DC circuit having a nominal voltage is proposed, the device comprising:

- an electrical energy storage element defined by a DC operating voltage lower than the nominal voltage, - a charger configured to charge the storage element when the storage element is discharged and the device is subjected to the nominal DC voltage, and

- a set of diodes defined by an associated threshold voltage value, the set of diodes being configured to be on and thus discharge the storage element when said storage element is charged and the device is subjected to a DC voltage lower than the difference between the DC operating voltage associated with the storage element and the threshold voltage value associated with the diode assembly.

[0020] If we consider a stationary nominal state, in which the device is subjected to a nominal DC voltage maintained at the terminals of an AC/DC or DC/DC converter and in which the electrical energy storage element is previously charged, then the device behaves as a whole as a simple electrical conductor, the set of diodes being blocked.

In such a configuration, when an electrical fault occurs downstream of the device, on a DC circuit portion to be protected, then this electrical fault causes a voltage drop. As soon as the device is subjected to a DC voltage sufficiently lower than the nominal voltage, then the set of diodes switches to on-state. The electrical energy storage element then discharges and maintains at its terminals a voltage equal to its operating voltage (i.e. lower than the nominal voltage). One consequence is that the device provides additional short-circuit current which can contribute to accelerating the melting of a fuse located in the portion of the DC circuit to be protected. Thus, it is possible to ensure that the melting of such a fuse occurs before the melting of a critical component of expensive equipment such as a semiconductor switch within an AC/DC or DC/DC converter. DC powering the DC circuit.

[0022] After disappearance of the electrical fault and restoration of the nominal DC voltage, a transient state is automatically established within the device: the set of diodes is blocked and the charger recharges the storage element. Once the storage element has been charged, the device resumes its operation in steady nominal state. [0023] It is known that a voltage drop can also find its source in an electrical overload not resulting from a circuit current but rather from a high demand. Such a drop in voltage can also trigger the flip-flop of the set of diodes in on-state, the discharge of the storage element and, consequently, the melting of a fuse on the portion of the circuit to be protected.

[0024] The characteristics set out in the following paragraphs can optionally be implemented. They can be implemented independently of each other or in combination with each other:

[0025] The storage element can be a capacitor or a supercapacitor, with the advantage of a particularly short response time, compatible with the requirements linked to the prevention of the destruction of semiconductor switches within a converter supplying the device.

[0026] Furthermore, supercapacitors have the advantage of reduced calendar aging compared to other types of storage elements, hence an increased lifetime. This is particularly true because, within the device, the storage element is more subject to calendar aging rather than aging by cycling.

The storage element can alternatively be a battery.

[0028] The charger may be of the step-down chopper type, or may, alternatively, comprise a resistive element and a semiconductor switch controlling the charging of the capacitor through the resistive element.

The set of diodes can be formed from a plurality of diodes connected in series.

[0030] Optionally, the device may comprise a plurality of disconnectors arranged so as to electrically isolate the storage element when the disconnectors are placed in the open position.

[0031] Optionally, the device may comprise a plurality of disconnectors arranged so as to electrically isolate the device when the disconnectors are placed in the open position. [0032] These disconnectors allow, by electrically isolating the potential sources of current passing through one or more branches within the device, an operator to protect himself upstream of maintenance operations on the device or on a circuit in which the device is installed.

According to another aspect of the invention, a system is proposed comprising:

- a converter interfacing a first circuit and a second DC circuit,

- at least one device-fuse assembly arranged within the second circuit, the assembly being formed of the above device and at least one fuse defined by a fuse rating,

- the device being arranged so as to, in the event of a short-circuit in the second circuit, discharge the storage element and thus supply the fuse with a first additional short-circuit current, and

- the converter comprising at least one capacitor configured to, in the event of a short-circuit in the second circuit, supply the fuse with a second additional short-circuit current, and

- the caliber of the fuse being chosen to cause, in the event of a short-circuit in the second circuit, a rupture of the fuse under the combined effect of a short-circuit current induced by the short-circuit, of the first supplement and of the second supplement before the converter is damaged.

The expressions “first circuit” and “second circuit” refer to the networks interfaced by the converter. The "first circuit" thus designates the network upstream of the converter, which is either alternating current (AC) or direct current (DC). Figure 1 represents for example, for this purpose, an AC network upstream of such a converter. The “second circuit” designates the network downstream of the converter which, in the context of the invention, is in direct current (DC).

The system may optionally comprise a plurality of device-fuse sets, said sets differing in at least one parameter chosen from the following list:

- the fuse rating,

- the DC operating voltage associated with the storage element of the device, And

- the threshold voltage value associated with the set of diodes of the device.

[0036] Thus, depending on the location of an electrical fault within the DC circuit, it is possible to select a device-fuse assembly that triggers first, in the sense that such a trigger consists of the supply of the first supplement short circuit by the device of the selected set, causing the melting of a fuse of the selected set.

Brief description of the drawings

[0037] Other characteristics, details and advantages will appear on reading the detailed description below, and on analyzing the appended drawings, in which:

Fig. 1

[0038][Fig. 1 ] is a circuit diagram of an AC/DC converter supplying a DC feeder protected by DC fuses.

Fig. 2

[0039] [Fig. 2] compares, in the context of an electrical fault occurring within a DC circuit as represented in FIG. 1, the melting times of 63 A fuses with those of semiconductor switches present in AC/DC converters. DC, depending on the distance between the position of the electrical fault and that of the converter.

Fig. 3

[0040] [Fig. 3] represents a Ragone diagram of different types of electrical energy storage elements.

Fig. 4

[0041 ][Fig. 4] is a diagram of a basic electronic circuit breaker, showing the current path in the event of an electrical fault.

Fig. 5

[0042] [Fig. 5] is a diagram of a more evolved electronic circuit breaker, on which the path of the current in the nominal regime is represented. Fig. 6

[0043] [Fig. 6] is a diagram of the electronic circuit breaker illustrated in FIG. 5, on which is also represented the path of the current in the event of an electrical fault.

Fig. 7

[0044] [Fig. 7] compares, in the context of an electrical fault occurring within a DC circuit as shown in Figure 1, the melting times of different types of fuses with those of semiconductor switches present in AC converters /DC oversized by 100%, depending on the distance between the position of the electrical fault and that of the converter.

Fig. 8

[0045][Fig. 8] represents a device according to an exemplary embodiment of the invention, which can be connected to a direct current network in parallel.

Fig. 9

[0046] [Fig. 9] represents a device according to an alternative embodiment of the invention.

Fig. 10

[0047][Fig. 10] represents the behavior of the device in figure 8 when it is subjected to nominal operating conditions, i.e. when there is no electrical fault in a downstream portion, to be protected, of a DC circuit.

Fig. 11

[0048][Fig. 11 ] represents the behavior of the device of figure 8 when it is subjected to a fault state, i.e. in the presence of an electrical fault in a downstream portion, to be protected, of a DC circuit.

Fig. 12

[0049][Fig. 12] represents the behavior of the device of FIG. 8 having previously been subjected to a fault condition, after disappearance of the fault.

Fig. 13 [0050][Fig. 13] is a diagram of a circuit of an AC/DC converter supplying a DC output protected by DC fuses, the circuit being equipped with a plurality of devices according to an exemplary embodiment.

Fig. 14

[0051 ][Fig. 14] compares, in the context of an electrical fault occurring within a DC circuit as shown in Figure 8, the melting times of different types of fuses with those of semiconductor switches present in AC converters /DC oversized by 100%, depending on the distance between the position of the electrical fault and that of the converter.

Fig. 15

[0052][Fig. 15] represents a variant of the device of FIG. 8, adapted to facilitate maintenance operations.

Fig. 16

[0053][Fig. 16] represents a variant of the device of FIG. 9, adapted to facilitate maintenance operations.

Description of embodiments

Reference is made to Figure 8 which shows an example of a device (50) according to the invention. Such a device is adapted to be connected to a direct current (DC) network in parallel.

[0055] Figure 8 highlights, by way of illustration, an example of such a direct current network in parallel. Are represented :

- an AC/DC or DC/DC converter (1) supplying a DC bus,

- a first example of a device according to the invention (50), positioned on the DC bus at the line head,

- a second example of a device according to the invention (60), positioned downstream of the first device (50),

- a plurality of fuses (5), arranged so as to form, respectively, with one of the devices (50, 60) above, a device-fuse assembly,

- a cable (6) represented as a resistive load, and - ultimately, a device (7) or group of devices to be electrically powered, represented in the form of a resistive load.

The first device-fuse assembly comprises the first example of a device (50) associated with at least one fuse (5), such a first assembly being responsible for protecting the downstream portion of the DC circuit denoted N, that is that is to say of the portion located more at the end of the line than is the device-fuse assembly (50) in question. The second device-fuse assembly comprises the second example of a device (60) associated with at least one fuse (5), such a second assembly being responsible for protecting the downstream portion of the DC circuit denoted N+1, that is that is to say of the portion located more at the end of the line than is the device-fuse assembly (50) in question.

The device of FIG. 8 comprises an electrical energy storage element (52), for example a capacitor, a supercapacitor or a battery, operating at a DC voltage, denoted O VDC, lower than the nominal voltage of the network DC, denoted VDC. In other words, the coefficient a has a positive fixed value and strictly less than 1. The voltage of the DC network is not strictly fixed but is on the contrary likely to vary over time, for example by 5% or 10% around the nominal value. To prevent such usual DC voltage variations from causing the device to trip, the coefficient a can be set, for example, to a value less than 0.9, or less than 0.85.

0n can also define a coefficient denoted y corresponding to the state of charge of the energy storage element. The coefficient y is equal to 1 (or 100%) when the storage element is fully charged, and equal to 0 (or 0%) when the storage element is fully discharged. Thus, in general, V storage is equal to y a-Voc.

To choose the most appropriate storage element to minimize the response time of the device, it is possible to refer for example to FIG. 3, which is a diagram comparing, for the different types of storage elements of electrical energy which are the supercapacitors (20), the LiC capacitors (21) and the lithium-ion batteries (22), the ratio between their maximum power density and their energy density. Capacitor-based technologies and supercapacitors are here preferred according to FIG. 3 because they present the fastest response.

The device of FIG. 8 also comprises a loader of the type of a buck chopper (51), also called step-down, arranged so as to be able to load the storage element (52). This converter only operates when the storage element has previously been discharged following an electrical fault: the converter then recharges the storage element. The buck chopper (51) can be replaced by a resistive charger (56) controlled by a semiconductor switch (55) (thyristor or transistor), as shown in Figure 9.

[0061] Finally, the device of Figure 8 also comprises one or more diodes (54), forming a "set of diodes" (53). A diode can be defined by its threshold voltage below which it is blocked and beyond which it becomes conductive. The set of diodes therefore has a threshold voltage value which can be determined from the threshold voltages of the diodes which form it.

In the device example of Figure 8, the set of diodes (53) is formed of KN identical diodes arranged in series. Such a set of diodes (53) has a threshold voltage value equal to the sum of the threshold voltages of the diodes which form it, that is to say KN times the threshold voltage Vs of a single diode. The set of diodes (53) allows the discharge of the storage element (52) if the potential difference between the bus and the storage element exceeds the threshold voltage value.

In nominal operation, the device of Figure 8 is in stand-by, as shown in Figure 10.

< o: your diodes are blocked

The charged storage element is at a voltage V standby storage equal to its operating voltage O VDC, therefore lower than the voltage VDC of the DC network. The charger does not conduct any current, ie the current \charger at the output of the charger is equal to zero. The diodes of the diode assembly are located in reverse bias, the diode voltage across the terminals of the set of diodes being less than zero, therefore necessarily less than the (strictly positive) threshold voltage value.

[0064] When a fault occurs, the DC bus voltage drops, taking a value Vcc lower than the nominal voltage VDC, which triggers an activation of the device, as shown in figure 11 .

Indeed, if the difference between the storage voltage V of the storage element and the voltage Vcc of the bus exceeds the voltage threshold value of the set of diodes, then the diodes become conductive. The storage element discharges, which brings an additional short-circuit current, denoted Ice boost, accelerating the melting of a fuse positioned downstream and thus providing faster network protection.

become drivers

fcc boost

Once the fault has been eliminated, the charge of the storage element is regenerated thanks to the charger as shown in figure 12.

[0066] Indeed, when the bus voltage is equal to the nominal voltage VDC of the DC circuit, then the set of diodes is blocked. The charging current \charger at the output of the charger also has a positive value as long as the storage element is at least partially discharged, that is to say as long as the coefficient y is less than 1.

diodes become blocked again fcarjtur ~~ ^DC early fcc

At the end of this phase, the storage element regains its initial charge, ie y=1, and a voltage equal to its operating standby storage voltage in nominal operation. The charging current \charger is canceled. The situation is then identical to that shown in Figure 10.

< o: the diodes are blocked

Reference is now made to FIG. 13 which represents a system comprising several examples of devices according to the invention.

An AC/DC or DC/DC converter supplies a DC output to which a DC load is connected. Between the power supply and the load, several fuse-device assemblies are arranged. Each device-fuse assembly comprises an example of a device according to the invention, associated with a fuse.

Figure 13 shows, by way of illustration, two fuse-device assemblies:

- a first device noted (N) associated with a first fuse is responsible for protecting a portion N of the DC circuit,

- a second device denoted (N+1) associated with a second fuse is responsible for protecting a sub-portion N+1 of the portion N.

These two device-fuse assemblies may be structurally different so that, when an electrical fault occurs in the sub-portion N+1, the device-fuse assembly responsible for protecting the sub-portion N+1 automatically triggers before the one responsible for protecting sub-portion N.

To do this, it is possible to influence various parameters. Thus, the device-fuse assemblies may differ:

- by the fuse rating,

- by the operating voltage of the energy storage element, and/or

- by the threshold voltage value of the set of diodes.

It is advantageous to provide that the device-fuse assemblies only differ from each other at the level of a single one of the aforementioned parameters.

[0073] For example, with a view to production on an industrial scale, the same generic base such as the device without diodes can be optionally supplemented by different versions of the diode assembly, each consisting of a different number of diodes in series. This makes it possible to be able to offer different versions of the device, respectively adapted to different needs, from a single and same generic base.

[0074] A fault as shown in FIG. 13 results in a more abrupt voltage drop at the level of the device protecting the sub-portion N+1 than at the level of the device protecting the portion N due to the impedance of the cable or of the the line connecting them. If one wishes to delay the reaction of the device protecting the portion N with respect to that of the device protecting the sub-portion N+1, one can choose the operating voltage of the storage element of the device protecting the portion N so that it is lower than the operating voltage of the storage element of the device protecting the sub-portion N+1.

[0075] Considering for example that the sets of diodes are each formed of diodes in series, it is also possible to play on the number of diodes in series in each of the devices in order to modulate the threshold voltage value to trigger the one of the devices before the other. Knowing that the threshold voltage of a diode is of the order of 1.4 V and that the voltage difference following a fault in two points of the DC network is low (a few V), by placing a reduced number of diodes in end-of-line devices compared to head-of-line devices, the equivalent threshold voltage of the diodes is lower at the end of line. Therefore, end-of-line devices bootstrap first.

Figure 14 shows the results of some simulations. The curves show the respective melting times of the semiconductor switches of the converter (70), of a 63 A DC fuse (71) and of an 80 A DC fuse (72) in the event of a DC fault according to its location on a feeder equipped with an example of a proposed device. In this example, the most critical location of the fault is located 10 m from the converter and the storage element has been sized to guarantee a minimum safety margin of 2 ms between the melting times of the fuses and that of the converter. As illustrated in Figures 15 and 16, all the examples of devices previously described can be equipped with disconnectors (57, 58, 59). The latter aim to allow electrical isolation of the device with respect to a DC circuit in which it is installed, and also aim, within the device itself, to electrically isolate the electrical energy storage element (52 ) to prevent accidental discharging.

The prior actuation of these disconnectors allows an operator to protect himself during an intervention on the device itself or more generally on the DC circuit connected to the device, for example during a maintenance operation.

More specifically, in Figures 15 and 16:

- two disconnectors (58, 59) are arranged to electrically isolate the storage element (52) when these two disconnectors are placed in the open position, and

- Two disconnectors (57, 58) are arranged to electrically isolate the device (50) from the DC circuit where it is installed when these two disconnectors are placed in the open position.

The various examples of devices according to the invention are intended mainly for public low-voltage direct current (LVDC) distribution networks and for uses supplied with LVDC, in particular industrial networks, LVDC, on-board LVDC networks, micro -networks, or autonomous direct current networks.

Non-patent literature

[0081] For all intents and purposes, the following non-patent elements are cited:

- nplcitl: K. HongJoo and al., "Development of high speed circuit breaker using IGBT drivers", KEPCO - Republic of Korea, 25th International Conference on Electricity Distribution, Madrid, 3-6 June 2019, Paper n° 1555;

- nplcit2: J. A. Giancaterino, "Dc-dc converter plants and their ability to clear distribution fuses", Proc. INTELEC, pp. 315-320, 1994;

- nplcit3: S. Ravyts, GV d. Broeck, L. Hallemans, MD Vecchia and J. Driesen, "Fuse-Based Short-Circuit Protection of Converter Controlled Low-Voltage DC Grids", IEEE Transactions on Power Electronics, vol. 35, no. 11, pp. 11694-11706,

Nov 2020.

Claims

[Claim 1] Device (50) capable of being inserted into a DC circuit having a nominal voltage, comprising:

- an electrical energy storage element (52) defined by a DC operating voltage lower than the nominal voltage,

- a charger (51, 55, 56) configured to charge the storage element when the storage element is discharged and the device is subjected to the nominal DC voltage, and

- a set of diodes (53) defined by an associated threshold voltage value, the set of diodes being configured to be on and thus discharge the storage element when said storage element is charged and the device is subjected to a voltage DC less than the difference between the DC operating voltage associated with the storage element and the threshold voltage value associated with the diode assembly.

[Claim 2] Device according to claim 1, in which the storage element is a capacitor or a supercapacitor.

[Claim 3] Device according to claim 1, in which the storage element is a battery.

[Claim 4] Device according to one of the preceding claims, in which the feeder is a step-down chopper (51).

[Claim 5] Device according to one of Claims 1 to 3, in which the charger comprises a resistive element (56) and a semiconductor switch (55) controlling the resistive element.

[Claim 6] Device according to one of the preceding claims, in which the diode assembly is formed by a plurality of diodes (54) connected in series.

[Claim 7] Device according to one of the preceding claims, comprising a plurality of disconnectors (58, 59) arranged so as to electrically isolate the storage element when the disconnectors are placed in the open position.

[Claim 8] Device according to one of the preceding claims, comprising a plurality of disconnectors (57, 58) arranged to electrically isolate the device when the disconnectors are placed in the open position.

[Claim 9] System comprising:

- a converter (1) interfacing a first circuit (2) and a second DC circuit (3),

- at least one device-fuse assembly arranged within the second circuit, the assembly being formed of a device (50) according to one of the preceding claims and of a fuse (5) defined by a fuse rating,

[Claim 10] System according to the preceding claim, comprising a plurality of device-fuse assemblies, said assemblies differing at the level of at least one parameter chosen from the following list:

- the fuse rating,

- the threshold voltage value associated with the set of diodes of the device.