Selective Signaling of Ground Faults in Networks with a Compensated Neutral
Abstract Generated abstract
This paper addresses selective signaling of ground faults in electrical networks with a compensated neutral, where arc-suppression coils reduce fault currents but changing network capacitance can disturb compensation. It argues that, despite undercompensation or overcompensation, ground faults produce zero-sequence current and voltage, with opposite initial current directions in damaged and undamaged feeders. On this basis, the paper proposes a selective signaling circuit using a highly sensitive power-direction relay fed by zero-sequence current and voltage filters. The relay employs phase-sensitive semiconductor amplification and a polarity-sensitive output relay to indicate whether the fault lies inside or outside the protected feeder, with possible additional amplification stages to improve sensitivity.
Full Text
Selective Ground-Fault Signaling in Networks with a Compensated Neutral
A. I. Zaitsev
(Presented by the scientific seminar of the Electromechanical Faculty)
The high rates of electrification of the national economy require the provision of maximally safe conditions for the use of electric power. One of the dangerous operating modes of an electrical network is a ground fault in networks with an isolated neutral. Sometimes such faults may be accompanied by an intermittent arc. The danger of an intermittent arc that arises consists in the appearance in the system of overvoltages reaching \(3.5\text{–}4.5\,U_{\phi}\). The possibility of an intermittent arc arising depends mainly on the magnitude of the ground-fault current. To eliminate the harmful consequences of an intermittent arc, the Rules for Electrical Installations provide for compensation of ground-fault currents by means of arc-suppression coils, which should be used in networks with voltages of \(20\text{–}35\ \mathrm{kV}\) when ground-fault currents exceed \(10\ \mathrm{A}\), and in networks with voltages of \(6\text{–}15\ \mathrm{kV}\) when currents exceed \(30\ \mathrm{A}\). With ideal tuning of the inductance, at the fault point there will flow the active component of the current, caused by active losses in the inductive coil, by leakage to ground due to imperfection of the conductor insulation, and by the higher harmonic components of the load current.
When the number of connected lines changes, compensation of the capacitive current may be disturbed, and as a result a considerable fault current may flow. Compensation may also be disturbed if the ground fault occurs at the beginning of a line, which, from the point of view of compensation effectiveness, will be equivalent to a reduction of the resultant capacitance by the amount of the capacitance of that line.
Therefore, in the absence of continuous adjustment of the inductance, undercompensation or overcompensation of the capacitive current may be produced. However, a common phenomenon at any degree of compensation is the presence, during ground faults, of a zero-sequence current and voltage. The directions of the zero-sequence currents in the damaged and undamaged lines in the first period, when the capacitances of the individual phases are being recharged, will be shifted by \(180^\circ\). Using this phenomenon, it is possible to create signaling for ground faults in networks with a compensated neutral.
The proposed selective signaling circuit (Fig. 1) responds to the direction and magnitude of the active component of the zero-sequence power flowing through the protected feeder in the first period from the moment of the ground fault. For this purpose, a highly sensitive power-direction relay is used. The relay circuit consists of a phase-sensitive power amplifier assembled on semiconductor elements and a high-speed output relay of the RP type, with transfer to self-holding when it operates. For ground-fault signaling in networks with a compensated neutral, the power-direction relay must be fed from zero-sequence current and voltage filters and must have maximum sensitivity at an angle equal to zero.
Fig. 1.
Amplification of the power taken from the zero-sequence current and voltage filters is carried out by a phase-sensitive rectifying device, in which a planar semiconductor triode connected in a common-emitter circuit is used as the amplifying stage. Depending on the phase relationship of the zero-sequence current and voltage, a direct current of one direction or the other will flow through the rectifier, and also through the winding of the output relay. The magnitude of the current flowing through the relay is determined by the values of the voltage and the active component of the zero-sequence current. The direction of the current flowing through the relay changes by \(180^\circ\) for a fault within and outside the protection zone. For selective operation, the output relay must respond to the polarity of the current. As such relays one may use a polarized relay of the RP-5 type, or another sensitive relay with its winding shunted by a diode in the direction opposite to that when the relay operates.
To increase the sensitivity of the power-direction relay, several intermediate stages of zero-sequence power amplification may be used.
Received by the editors
in June 1962.