Earthfault:

Earthfault detection in insulated medium-voltage network must be designed properly. Two methods may be used to detect an earthfault:

Both principles may be combined to protect a power network:








An earthfault in an insulated network leads to an unsymmetrical voltage system (U30~0 in our example) and therefore a displacement voltage

.

So existence of an earthfault is simply proved (ANSI 59N), the location is not.

The sum of currents through the phase-to-earth capacitance becomes different from zero:

A current is flowing through the point of earthfault and the cable capacitance (blue line) using the ships hull. The zero-sequence current flows through a ZCT connected to a device with protection function 51N. So the place of earthfault can be identified by metering the ZCTs secondary current.

Often a high-resistance grounding (Resistor RE at generator G1´s starpoint) is used to improve the power factor at the place of earthfault. The high-resistance grounding leads to an additional ground current (red line). This ground current is flowing through the generator notwithstanding the earthfaults location. An differential mechanism (87N) is used to identify inner faults of the generator. This mechanism is restricted to the protection zone between the current transformers!

If network is not designed for permanent operation with an earthfault (Resistor RE not designed for continous load!!), a protection scheme must be designed to isolate the fault in three steps:

  1. The faulty device should be disconnected by protection function 51N or 87N with short delay.(e.g. 2 s)

  2. If the main switchboard can be divided by a tiebreaker, the tiebreaker should be opened by ANSI 59N, to limit the earthfaults effects to one one side. (And to limit the blackout to one side, if step 3 takes effect)

  3. If the earthfaults location can be not be identified, all working generators must be disconnected by ANSI 59N to protect the resistor etc.




UH,27.10.04,16:24:04