SE1000829A1 - Shuntanordning - Google Patents

Abstract

Uppfinningen hanfor sig till en shuntanordning (1) for skydd av en omvandlare i ett elektriskt kraftoverforings-och/eller distributionsnatverk, varvid shuntanordningen (1) ar parallellkopplad med en omvandlarsubenhet (2) i omvandlaren. Shuntanordningen (1) innefattar ett forsta delnatverk (la) och ett andra delnatverk (lb), varvid de forsta och andra delnatverken (la, lb) ar identiska och innefattar en respektive tyristor (4, 5) . De forsta och andra delnatverken (la, lb) ar parallellkopplade med sina polariteter omvanda, och varje delnatverk (la, lb) ar anordnat att tillhandahalla en shuntvag for omvandlarsubenheten vid funktionsavbrott darav for ett stromtecken i enlighet med en riktning av dess tyristor (4, 5) .(Fig. 1)

Description

20 25 cells in SVC (static VAr compensator) converters or even single switches of HVDC phase legs.

A voltage contribution of one such sub-unit to the overall is by design choice the voltages synthesized by the converter, and thus purposely, not essential for converter operation. A reduced number of composing sub~units may, and always should be able to, fail while still behaving like voltage sources not substantially exceeding their rated values. This clearly includes the case of failures such as short-circuit, which is a particular case of voltage source.

Conversely, if even one sub-unit would be allowed to fail as a current source, the consequent constraint imposed on the grid current might lead to inhibition of the converter and the interruption of the service. The open circuit fault, even. just in one direction of the current, is clearly a particular case of this scenario.

Summary of the invention In view of the above, it is a general object of the invention to provide means enabling the uninterrupted grid operation even in case of component failures of a converter of the grid.

It is particular object of the invention to provide means for enabling uninterrupted operation of a converter even in case of component failures thereof.

These objects, among others, are achieved by a shunt device as claimed in the attached independent claims.

In accordance with the invention a shunt device for an electric power transmission and/or distribution system is 10 15 20 25 provided. The shunt device is connected in parallel with a converter sub-unit of the converter and comprises a first sub-network and a second sub~network, the first and second sub-networks being identical and comprising a respective thyristor. The first and second sub-networks are connected and each sub- the in parallel with their polarities reversed, network is arranged to provide a shunting path to converter sub-unit upon failure thereof for a current sign according to a direction of its thyristor. By means of the invention, means are provided for enabling continuity of grid operation even' during failure of a sub-unit of a converter of the grid. The invention provides an alternative current path automatically upon detection of an alteration of a normal current path. The shunt device is most reliable in that no sensors or complex controller are used. Further, the shunt device of the invention avoids the use of external energy supplies for its operations, as such external energy supplies might be destroyed as well by the fault. Yet again, improved reliability is thereby provided.

Further features and advantages thereof will become clear upon reading the following description and the accompanying drawings.

Brief description of the drawings Figure l illustrates an embodiment of the present invention.

Detailed description As mentioned in the background section, although a converter the function of the converter should be should behave like sub-unit fails, maintained. The converter sub-units normal voltages sources, not exceeding their rating, even 10 15 20 25 during failures such as short-circuits or open circuit faults.

It is therefore necessary to detect the alteration of the normal current path caused by the appearance of a “current- source” behavior and to provide automatically, yet very rapidly, an alternative “shunting” bidirectional current path. This guarantees that the grid current can continue to flow despite of type of damage that has affected the sub- unit which from this event onwards can be kept inert, then serviced at the earliest convenience.

The present invention provides a two-terminal autonomous shunt device capable of behaving as an automatic and permanent shunt, should some composing converter sub-unit fail as an open circuit or be forcedly excluded from contributing to the current path (like in case of detected arcing inside faulty semiconductor devices). The shunt can also intervene to suppress independently arising overvoltage conditions across the converter sub-unit.

The shunt device in accordance with the invention restores and maintains an alternative path for the grid current until the converter is repaired. For its operation, the shunt device relies on passive components and rugged semiconductors (such as thyristors) only, without any use of sensors or complex controllers that would just reduce its reliability. For this same reason the shunt device does not depend on external energy supplies for its operation, supplies which might be destroyed as well by the fault in the converter sub-unit. 10 15 20 25 The automatic shunt device in accordance with the invention is an independent two-terminal device that does not require modifications of existing electrical designs. The shunt device could be added. to already existing installations, provided that næchanical and cooling system compatibility exists.

The capability of providing an alternative current path automatically, thereby increasing the availability of the converter and the continuity of the grid operation, provides economical advantages as well as safety advantages.

Figure l illustrates an embodiment of the present invention.

In particular, in the illustrated embodiment, a two-terminal shunt device achieves the above desired functionality.

The converter and the grid to which the converter is connected appear as a current source to the faulty converter sub-unit. This is correct provided that the sub-unit fails “acceptably”, i.e. as a voltage source of proper magnitude, thereby without forcedly altering the grid current. Under this assumption everything external to the faulty sub~unit can be modeled as a current source Igrid illustrated at the left hand side of figure 1. It is however to be noted that no hypothesis on the waveform of this current is necessary for the operation of the shunt device.

The two-terminal shunt device 1. is connected directly in parallel with a sub-unit which could fail as a current a converter sub-unit 2 such as a the source. .As an. example, switching device of the converter, could be mentioned, failure of which could pose constraints on the converter current. The converter sub-unit 2 thus represents one two- 10 15 20 25 terminal sub-unit that can fail as a current source. The currents of the converter sub-unit 2 and of the shunt device l are indicated with Iunit and Is, respectively, whereas the voltage common to the converter sub-unit 2, the shunt device l and the generator Igrid is denoted vs.

The shunt device 1 comprises two identical sub-networks la and lb, and could be considered as comprising two “sides”, denoted “Positiv sida” and “Negativ sida", respectively, in figure l. Each side, a “positive” side corresponding to a first sub-network la and a “negative” side lb corresponding to a second sub-network lb, is devoted to one specific sign, namely + or - of the current IS. Each sub-network acts to provide a shunting path for the current sign according to 5 may (SCR). the direction of its thyristor 4, 5. The thyristors 4, for example be a semi-conductor controlled rectifier The two sub-networks la, lb connected in “anti-parallel", i.e. connected in parallel with their polarities reversed, provide the necessary bidirectional shunting capability.

Obviously Is = Igrid - Iunit holds true at any instant and during any condition.

Igrid is the current of a current source 3 nmdeling the behavior of the remaining converter units (not illustrated) and the grid (not illustrated) to which the converter comprising the converter units is connected.

Is is the current to the shunt device l, and Iunit is the current to the converter sub-unit 2.

The key components that radically change the behavior of the shunt device 1 in case of abnormal operation of the converter sub~unit 2 are two fuses Fp and Fn. The fuses Fp, 10 15 20 25 Fn. are forced to intervene in case of a fault and the circuit behavior will be permanently different after such event.

It is assumed that both fuses Fp (Fn) are healthy in normal operating conditions of the converter sub-unit 2, thereby and the whole (Zln, the short-circuiting the capacitors C2p (C2n) RGp, iGp, RGn, iGn, devoted to group of components Zlp, Z2n, Z2p, Lsp Lsn) which are ignite “positive" (“negative”) thyristor 4 (5). Because of the nature of the components involved, in such a condition the whole current Is flows through the fuses Fp and Fn as well as the capacitors Clp and Cln. The presence of the diodes Dp and Dn assure that the “positive" (“negative”) side of the shunt device 1, i.e. first sub-network la and second sub-network lb, respectively, is interested by the positive (negative) values of Is only. Dp and Dn are “transparent” to the sub- systems outside the shunt device l itself.

Until the I2t integrals of both the positive and negative values of Is remain below the I2 rating of the fuses Fp and Fn, such fuses do not intervene and the shunt device l appears to be outside the circuit as a capacitance Cl = C2.

The capacitance Cl, C2 is chosen suitably small.

Under these conditions a very good approximation is: Is = Igrid - Iunit = Clp*dVs/dt, wherein Vs is the voltage over the converter sub~unit 2.

Clearly, the breakdown voltage of diodes TXp (TZn) needs to be greater than the maximum Vs in normal operating conditions. 10 15 20 25 In the following the functioning of the shunt device is described.

I) First detection of abnormal current in failing converter sub-unit and subsequent intervention of fuse(s) When an abnormal condition inside the converter sub-unit 2 modifier Iunit to such an extent that the Izt integral of Is the = Igrid ~ Iunit exceeds the rating of the fuses Fp (Fn), (Fn) side of the RGn, (negative) Indeed, (Z1n, fuse Fp intervenes, and the positive the shunt device l is drastically modified. whole group of components Zlp, RGp, iGp, Z2p, Lsp iGn, Z2n, Lsn) devoted to ignite the “positive" (“negative") thyristor 4, 5 now become active. An example of such abnormal conditions inside the converter sub-unit 2 is an open-circuit fault, even unidirectional only.

The sudden open circuit fault in the converter sub-unit 2 is a particular case of this scenario. It is the case indeed where Iunit changes abruptly from its normal value at that specific instant, to zero.

It is noted that diodes Dp and Dn do not need to be rated for the whole Igrid current in continuous operation. It is sufficient that their short duration peak current is equal to the maximum of Igrid. When the thyristors 4, 5 enter full conduction, the current Is transfers to them.

The increase in the Izt integral of Is can be obtained also indirectly by the raise of the voltage Vs, when it exceeds the breakdown of diode TZp. This might occur independently from substantial changes in the converter sub-unit current and it constitutes an independent overvoltage protection capability of the shunt device 1 in accordance with the 10 15 20 25 invention. in such a case the current surge in TZp is the one that contributes significantly to the intervention of the fuse.

It is also noted that if the current Is was DC and, in addition, if it was incapable of creating the intervention of the fuses due to its too low value, it would flow in the capacitor Clp (Cln) until its voltage reached the breakdown of TZp (TZn). From such instant and onwards, the current would flow in Fp (Fn) and TZp (TZn) with Vs clamped at the breakdown voltage. Also in this case, a shunting path for Igrid would therefore exist.

The circuit behavior is in the following described only with reference to the “positive” side of the shunt device l, relevant for Is = Igrid - Iunit > O. The “negative” side evolves similarly when Is = Igrid - Iunit < 0.

After the intervention by the shunt device l, and thus opening of fuse Fp, the following sequence of events occur: 1) The current Is > O, previously flowing in Fp, is now diverted into capacitor C2p which raises its voltage starting from zero initial condition. The capacitance C2p, like Clp, small to provide a It is must be sufficiently conveniently rapid operation of the shunt device 1. noted that the capacitors Clp and C2p should be implemented using low-inductance, capacitors snubber-grade apt to withstand current surges of short duration, equal to the maximum of Igrid. 2) When the voltage across C2p exceeds the breakdown voltage of Zlp, the current begins to flow in the gate of the positive thyristor 4. Its value can ignite the thyristor 4 10 15 20 25 10 when the Gate-Cathode firing characteristics is properly biased.

The gate triggering circuit must assure that the gate is excited by the proper current pulse, but at the same time it must guarantee that an additional current path always exists to avoid that the whole current Is, which potentially could equal Igrid if Iunit = O, flow through the gate terminal, even for a few micro seconds, before the thyristor 4 enters full conduction. This is a key reason as to why the capacitor C2p and transient voltage suppression (TVS) diode Z2p should used. Allowing the whole current Is into the gate, even for such a short duration of time, could be very harmful for the thyristor 4. not contradictorily, in a (Zln) Nevertheless, and preferred embodiment of the invention, are chosen to be a DIAC Zlp (diode for alternating current) in order to provide a proper current impulse to the gate of the thyristor 4. In such a case the use of Z2p is highly preferred too, in order to clamp the voltage feeding the gate circuit, thereby limiting the peak gate current to approximately: IGp_peak = (Vz2p ~ Vdiac - Vgk)/RGp, where Vdiac is the stable DIAC voltage and Vgk the Gate- Cathode voltage according to the Gate-Cathode firing Characteristics.

It is noted that Z2p - and Zlp if a DIAC is not used - are preferably not simple low-signal Zener or Avalanche diodes.

Instead they are preferably common transient suppression zener diodes (transorbs). Oppositely from conventional Zener diodes, the transorbs can withstand high peak currents of 10 15 20 25 ll short duration and are guaranteed to fail firstly as short circuits, should this happen. Such a consideration is particularly important for Z2p, since after it clamps the voltage, most of the current Is flows inside it until the thyristor 4 enters a sustained conduction state.

It is further noted that if the current capability of Z2p were still of concern, even before the thyristor 4 enters full conduction state, a saturable inductor (or reactor) Lsp could be inserted in parallel to it in order to divert the current from the diode. The inductor Lsp should be designed with a voltage-integral rating that brings it in heavy saturation after the voltage has been clamped by Z2p for a desired time interval. Further design criteria for the saturable reactor Lsp must consider the low-frequency zero it provides for the voltage across C2p. This can hinder the operation of the circuit if Is had only a very low frequency (almost DC), content since the reactor Lsp is in parallel to the fuse Fp. 3) When the thyristor 4 enters complete conduction, the current Is (> O) is completely diverted through it, thereby creating the shunting current path. The capacitor C2p has been discharged by the gate circuit, whereas Clp might remain charged to a voltage never exceeding the breakdown voltage of TZp.

It is noted that it is advisable to connect a resistor in parallel to CIO in order to discharge it before the next intervention of the positive side of the shunt device 1. 4) The positive side of the shunt device 1 ceases the conduction when. Is > O becomes smaller than the holding 10 15 20 25 12 current of the positive thyristor 4. Clp and C2p act also as snubbers, i.e. devices suppressing voltage transients, for the commutation of the thyristor 4.

II) Subsequent and continued operation with open fuse(s) With the fuses Fp (Fn) now open, as a consequence of a previous fault detection, the thyristors 4, 5 become ignited already at low values of Is. This is true provided that Clp and C2p are conveniently small, as previously noted. When Is the current will flow immediately into C2p the Z2p. The > 0 again occurs, until its voltage reaches breakdown of thyristor trigger Circuit operates as described under I) but since C2p is small, the breakdown voltage Z2p and the breakdown voltage of DIAC Zlp will be reached very rapidly.

As a consequence, the positive thyristor 4 will be ignited thereby providing the shunting path once again, and so already at low valued of Is.

The rapid creation of the current path provided by the thyristor 4, 5 is a consequence of the absence of short- circuit across the capacitor C2p. This is a key topological modification introduced by the intervention of the fuse Fp, which otherwise impedes the voltage raise across C2p in normal operating conditions, thereby inhibiting the whole thyristor trigger circuit.

The alternate intervention of the “positive” and “negative” side of the shunt device l (if Igrid assumes both signs) continues indefinitely until the fuses Fp and Fn are replaced, following the restoration of the failed converter sub-unit 2 to its normal functionality. 10 15 20 25 13 It is observed that once the failed converter sub-unit 2 is functional again and the shunt device is no longer needed, the replacement of Fp and Fn can be performed “on the fly” without the necessity of interrupting the operation of the converter. As soon as new fuses are inserted, the trigger circuits are inhibited and the thyristors will not ignite any longer starting from the next half-cycle.

Due to the automatic intervention of the shunt device 1, the failed converter sub-unit 2can. be removed and reinserted while the converter is in operation. A still functional converter sub-unit could also be removed in principle, since its disconnection can be “seen” simply as an open circuit fault by the shunt device l.

The shunt rated current and voltage are linked to the and the shunt device 1 is therefore with thyristors 4, 5 used, easily scalable to different ratings. All components, obvious exclusion of the thyristors 4, 5 do not need to be large. A cost-efficient device is thus provided.

An external firing command for each thyristor 4, 5 of the shunt device 1, e.g. a command possibly originating from the converter controller or a “manual" emergency operation, can be easily provided by adding a branch to the thyristor trigger circuit.

High values of dVs/dt can clearly lead to currents in Fp and Fn capable' of providing their intervention. It descends that, by proper design of Clp and Cln, the shunt device l can acquire also functionality of protection against dangerous dV/dt, should this be a concern for the convert sub-units. 14 It is noted that in other embodiments of the invention, the inductors Lsp, Lsn may be omitted, as they are not strictly necessary in all cases.

It is further noted that in other embodiments of the invention, the diodes TZp, TZn, e.g. Zener diodes, may be omitted, as they are not necessary in all cases.

Claims (7)

10 15 20 25 15 Claims

1. A shunt device (l) for protection of an apparatus being a part of an electric power transmission and/or distribution said shunt device (l) network, being connected in parallel with an apparatus sub-unit (2) of said apparatus, characterized in that said shunt device (l) (la) first and second sub-networks comprises a and a second sub-network (lb), said (la, lb) first sub-network being identical and comprising' a respective thyristor (4, 5), said first and second. sub-networks (la, lb) being connected. in parallel with their polarities reversed, each sub-network (la, lb) being arranged to provide a shunting path to said apparatus sub-unit (2) upon failure thereof for a current sign according to a direction of its thyristor (4, 5).

2. The shunt device (l) wherein each (la, lb) as claimed in claim l, sub-network comprises a fuse (Fp, Fn) arranged to intervene in case of a fault of said apparatus sub~unit (2).

3. The shunt device (l) as claimed in claim l or 2, wherein each sub-network (la, lb) comprises means (Zlp, RGp, iGp, Z2p, Lsp; Zln, RGn, iGn, Z2n, Lsn) arranged to ignite its thyristor (4, 5).

4. The shunt device (l) as claimed in any of the preceding claims, wherein each sub-network (la, lb) comprises means (Dp, Dn) arranged to allow only a current of a particular sign to flow therein.

5. The shunt device (l) as claimed in any of the preceding claims, wherein each sub-network (la, lb) comprises means 16 (C2p, Z2p; C2n, Z2n) arranged to limit the current into its thyristor (4, 5) gate.

6. The shunt device (1) as claimed in any of the preceding claims, wherein each sub-network (la, lb) comprises means (Clp, Fp; Cln, Fn) arranged to provide a current path during normal operation of said apparatus sub-unit (2).

7. The shunt device (1) as claimed in any of the preceding claims, wherein said apparatus is a converter.