The invention concerns a device for earth fault current

com pensation, com prising two or more fast adjustable reactance elements, arranged to com pensate for an earth fau lt current in a power network with a power transformer according to claim 1 .

Furthermore the invention concerns a method to create a neutral voltage in a power network with a power transformer according to claim 7.

Background of the invention and prior art

Today, transfer of electric power is mainly accom plished by means of cyclo-sym metrical three phase voltage systems. To fulfil the basic demands of the system - production and consum ption of power m ust be in balance at all times. In order to facilitate th is requirement large national transm ission networks have been created, where all producers and consumers of electric power are interconnected. The connection of these national grids to transnational networks provides further advantages with respect to the basic balance requirement. One such transnational network for exam plee is the Scandinavian NORD EL network.

To reduce the losses in the transm ission networks, the long distance transport of energy is accomplished at high voltage, preferably

400kV. The energy output to the consumers from the transm ission network is carried out via grid exit transformers, which in turn, supply a lim ited geographical area. The distribution is mainly effected at 1 0- 30kV voltage levels. Large industry customers are connected directly to the distribution network, while smaller consumers and households are suppl ied over yet another transformation to 400/230V.

Due to their meshed structure, transm ission networks have relatively high availability. Distribution networks, however, normally have radial structures, and therefore constitute a weak link in the overall power transm ission system . Faults on individual distribution lines may lead to disconnection of large groups of consumers. Distribution l ine fault protection therefor m ust not only ensure highest possible public safety and fire protection but at the same time also keep the

distribution networks available. At a first glance these two aims look difficult to combine.

But, in this context it is important to understand that the majority of the line faults are caused by single phase-to-earth insulation break downs - so called earth faults - while the energy transfer takes place between the phases. Earth does not take part in the transfer of energy. From a protection point of view it would be desirable to treat earth faults separately, and if possible choke the fault current in a systematic way down to levels where an im mediate line

disconnection can be avoided.

The most successful protection concept based on this understanding was developed in 1 91 7 by Waldemar Petersen. By connecting the neutral of the power system to earth over an adjustable inductance, wh ich is tuned to the grid capacitive leakage, Petersen could reduce the earth fault current by a factor of ten up to fifty. This reduction is usually enough to ensure the self-extinction of single phase flashover faults, which is the most frequent fault type on overhead lines.

Today, Petersen's resonance earthing dom inates in the Scandinavian and other European distribution networks. Due to the self-extinction of flashover faults, these networks have an overal l availability that is superior to other com parable distribution networks, employing other concepts for system earthing. In the ongoing conversion of the distribution networks from overhead line to underground cable networks the capacitive leakage currents increase dramatically by a factor 30 to 50 due to the m uch higher specific capacitance of cables. This affects Petersen's resonance earthing twofold: On one hand the self-quenching effect on flashover faults in the remaining overhead grid decreases successively and finally stops working due to increasing residual fault currents which are not com pensated by Petersen's "arc suppression coil". On the other hand the self-quenching effect does not work at all in cable grids, due to the short distance between live conductor and the earthed cable shield.

The problem was identified and finally solved in the beginning of the nineties. New, computer based technology for residual current com pensation (RCC) was developed to be used as a complement to the arc suppression coil of Waldemar Petersen. The new technology made resonance earthing again the superior concept for both overhead and cable networks. Today resonance earthing with RCC technology is used worldwide on voltage levels from 6kV to 1 1 0kV, now even in grids which previously em ployed other earthing concepts.

In contrast to the traditional arc suppression coil , which sim ply increases the source im pedance in the earth fault circuit, the RCC elim inates the driving voltage in that circuit by superim posing an equal but opposite voltage. This is not just a trivial task, since the exact driving voltage is unknown at the beginning of an earth fault.

According to Thevenin's theorem the fault current is determ ined by the driving voltage at the fault side and the im pedance of the fault in series with the source im pedance. Conversely, to elim inate the fault current com pletely, the driving voltage at the fault side must be elim inated.

The driving voltage at the fault side is made up by the phase-to-earth voltage of the supplying transformer (known, respectively

measurable at the substation), plus the load current dependent voltage drop between the supplying transformer and the actual fault side, which could be far out in the grid (unknown).

Determ ining the latter became possible by the development of a new com puter based algorithm , which was first published in the beginning of the nineties (Winter, K. "Swedish D istribution Networks - A New Methode for Earthfault Protection in Cable and Overhead Systems", 5th International Conference on Power System Protection, I EE conference publ ication no 368, York/U K 1 993). The remaining problem , to generate a voltage, possible to control with respect to am plitude and phase in relation to the supplying transformer was solved first by an arrangement of two phase shifting transformers. (Swedish patent application no S E437096, 1 984). As the mechanical control of a phase shifting transformer is relatively slow, this solution was later replaced by power electronic (pulse- width modulating inverters) Today reference installations for the com bined ASC/RCC with inverter technology are in operation worldwide in many power grids from 6kV up to 1 1 0kV. The necessary power for the RCC inverter is dependent on voltage level and grid size in terms of capacitive leakage current and dam ping factor. Large cable grids in the urban areas of the world may exceed 1 000A capacitive leakage current with an

uncom pensated residual fault current in the order of 1 00A or more. Expensive inverters, larger than 1 MVA, are required to compensate these huge currents.

The present invention propose a novel , more rel iable and more cost efficient arrangement for the com pensation of the residual earth fault current.

Summary of the invention

The objective of the invention is to solve the above problem s and offer a sim pler and more robust solution for earth fault current com pensation. A further objective is to com pensate the earth fault current com pletely and thereby im prove safety. Yet another objective is to im prove the reliabi lity of the com pensation device and thereby the rel iability of the whole power transm ission. Finally, to lower the costs for the earth fault current com pensation is also an objective of the invention. These objectives are achieved by means of a device, defined in the preamble of claim 1 , which is characterized by at least two

controllable reactance elements, each connected with one side to a phase of a three phase power source, wh ich is synchronous with the main power system , and where the other side of the reactance elements is jointly connected to the neutral of the main power system or its equivalent. In this way the neutral voltage of the main power system and its phase can be controlled by the size of the reactance elements, which in turn are controlled by a control device for the neutral voltage in order to elim inate the driving voltage for an earth fault current.

The advantage of the invention is im proved safety by a very fast elim ination of voltage injection and earth fault current into the fault side. No im mediate line disconnection is necessary for this. Therefor the new invention im proves not only publ ic safety and fire protection but also supply qual ity. The device is sim ple to manufacture and offers a low cost upgrade for existing arc suppression coi ls. Furthermore the device is suitable for different types of AC power networks like, two, three or other multi-phase systems.

According to another em bodiment of the invention the mentioned control device is connected to the reactance elements, for online control of phase-to-ground voltages in order to monitor grid

insulation online and allow the early detection of latent insulation issues like for instance bad surge arresters etc.

According to a further em bodiment of the invention, the control device will decide which phases to connect to the two reactance elements.

According to another em bodiment of the invention, the control device wil l control the size of the reactance elements until the conditions for the full earth fault current com pensation are fulfilled. The device according to the invention may thereby be used together with known control and measurement units for residual current com pensation (RCC). According to yet another em bodiment of the invention the device is arranged to work in parallel with an existing Petersen coil in order to just elim inate the residual fault current which is not com pensated by the Petersen coil. But in this arrangement the device can also be used j ust to increase the power of the Petersen coil .

These purposes are achieved by means of a method, defined in the pream ble of claim 7, which is characterized by the use of at least two adjustable reactance elements, each connected with one side to a phase of a three phase power source, which is synchronous with the main power system and where the other side of the reactance elements is jointly connected to the neutral of the main power system or its equivalent. In this way the neutral voltage of the main power system and its phase can be controlled by the size of the reactance elements, which in turn are control led by a device for the control of the neutral voltage in order to elim inate the driving voltage for an earth fault current. The method according to the invention im proves the rel iability of existing com pensation devices and thereby also im proves the availabil ity of the power supply.

The purpose is achieved by a method to generate a neutral voltage in a power network which is suppl ied from a power transformer, characterized by an arrangement including at least two controllable reactance elements, each connected with one side to a phase of a three phase power source, which is synchronous with the main power system and where the other side of the reactance elements is jointly connected to the neutral of the main power system or its equivalent, wh ile a control device is connected to the reactance elements in order to control the neutral voltage of the main system , and where:

- the control unit adj usts the neutral voltage with respect to

am plitude and phase angle in relation to the voltages of the main power transformer

- the control unit decides which reactance elements are

connected to the neutral of the main power system

- the control unit adjusts the size of the reactance elements unti l the conditions for the full com pensation of an earth fault current are met. According to one em bodiment of the method only two reactance elements are used, where the control unit first decides to which of the two phases of a synchronous three phase AC power source the elements are connected. According to another em bodiment of the method three reactance elements are used to speed up the fault interception by direct control of the concerned reactance elements, while the third element is left off. The purpose can be achieved likewise by use of one of the

arrangements of the method in parallel to an existing Petersen coil, just to take care of the residual fault current which is not

com pensated by the Petersen coil. Brief description of the drawings

Figure 1 shows a single line diagram of a distribution network with a known arrangement for the com pensation of an earth fault current

Figure 2 shows the same distribution network with an arrangement for the compensation of an earth fault current according to the invention. Figure 3 shows the vector diagram for the driving voltage dependent on the location of the earth fault in the distribution network. Figure 4a shows the detailed circuit diagram for an arrangement according to the invention.

Figure 4b shows the vector diagram for that arrangement. Detailed description of the invention

Figures 1 and 2 show a distribution network with a control and measurement unit 2 for the detection of an earth fault and the control of a fault current compensation device. Furthermore the figures show a power transformer 3, a busbar 4 and a number of outgoing feeders Li, l_2 ... LN, each with a line breaker 9 at the busbar 4.

Electric power transmission is generally effected by means of cyclo symmetrical three phase systems. However, even single and two phase systems (railway) exist. The common ground is that transfer of payload (to the consumers) is driven exclusively by the voltages between the phases. If an error occurs in form of an insulation break down between these phases, the feeder in question and all the consumers behind are disconnected by the line breaker 9. However, the majority of electrical faults occur between one of the phases and earth (so called earth faults). Thereby the entire system is shifted in relation to earth. Nevertheless, the voltages between the phases (which are driving the payload) are not affected. The fault current to earth at the actual fault side is determined by other currents to earth (mainly capacitive but also resistive leakage currents) in the galvanically interconnected network. The sum of all these currents to earth, including the fault current, is always zero (Kirchhoff's current law). From this follows the conclusion: If one wants to bring the fault current to zero, you must make sure all other currents to earth sum up to zero. The purpose of the com pensation devices in both figure 1 and 2 is to create this balance by producing a corresponding current between the neutral of the network and earth.

Figure 1 thereby shows a known device for this compensation, consisting of a Petersen coi l 5 with an auxiliary power winding 6 for an inverter 7 to inject the residual fault current, which is not com pensated by the Petersen coil. An option for back up tripping of the faulty feeder 8 in cause of a malfunction of the com pensation device is also shown.

Figure 2 shows an arrangement for com pensation of the complete fault current by two reactance elements according to the invention.

The control and measurement device 2 continuously monitors the zero sequence adm ittance (Yo) on the outgoing feeders L I -N and in the earth connection of the com pensation device. The actual values are stored in the memory of the device. Upon detection of an earth fault the measurement is repeated. Then the difference between the stored and last value is calculated feeder by feeder.

The fault is on the feeder which shows a difference (delta Yo) between the two values. The difference which also can be seen in the earth connection of the com pensation device is used to control the setting of the compensation device. In figure 1 it is the known arrangement with an inverter, while in figure 2 two adjustable reactance elements according to the invention are used to achieve balance. In both cases a controllable voltage is created between the neutral of the network N and earth E .

When this neutral voltage corresponds to the driving voltage at the fault side, delta Yo and thereby the current in the fault becomes zero (Thevenins teorem ). As Figure 3 shows a controllable neutral voltage is required to inhibit the driving voltage at the fault side. The compensation device has to create this voltage, which must be adj ustable in both ampl itude and phase with respect to the voltage system of the suppling power transformer.

Figure 4a shows a circuit diagram for the arrangement of two controllable reactance elements according to the invention. A reactance element is per definition a passive 2-pole, either

capacitive or inductive. In the example in figure 4a the reactance elements are made up of two tuneable capacitor banks, already known and used as fast tuning range for different types of fixed Petersen coils. The control unit first selects to which two phases the reactance elements have to be connected and then individually controls the size of the elements until the condition for full fault current

com pensation (delta Yo = 0 m S) are fulfilled. Figure 4b finally shows the vector diagram for the com pensation arrangement of Figure 4a. The delta connected primary winding of the source transformer induces on the wye-connected secondary side three cyclo sym metrical voltages with 1 20° phase shift. Together with the two adjustable reactance elements and the three possible permutations of two of these voltages it is possible to create a resulting voltage which can be varied 0-1 00% in am plitude and 1 20° in phase for each perm utation, i. e. 360° in total.

The device can be used for the com plete com pensation of the earth fault current as well as in paral lel to an existing Petersen coi l just for the com pensation of the residual fault current which is not

com pensated by the Petersen coil.

The invention is not lim ited to the above shown arrangement, but can be varied in many ways within the frame of the below following claims.