Detecting a ground fault on a Delta System 2

[CuriousElectron]

Hi,
Now days with a smart relay, can a single-phase ground fault be detected with digital relays without utilizing the typical ground fault scheme utilizing residual/zero-sequence CTs? If there is a single ground fault, there would be zero current flow in one of the phases. The relay will detect this single phasing event and will trip.

It sounds like the relay can be programmed to calculate this single ground fault on ungrounded system without elaborate zero-sequence CT/VT schemes.
Am I right on this thinking?
Thanks,
EE

 

[waross]

Normally insulation resistance and capacitive effects will keep delta voltages to ground at about 58% of line to line voltages.
A grounded phase will pull one phase to ground potential and raise the ungrounded phases to lint to line voltage to phase.
No real effect on the current.
The relay must have a ground reference to detect ground faults on an ungrounded circuit.
The first ground detection systems were three lamps connected line to ground.
In the event of a ground on one phase, that phase light would go dark and the other two would get brighter.
Later systems used relays in place of the lamps.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[OffTap100]

Using a ground bank with a broken delta connection, and a Delta breaker scheme, yes.

OT100

 

[FreddyNurk]

Detection of ground faults with an ungrounded delta scheme can be done via an insulation monitor (e.g. a Bender unit), or as OffTap100 suggested, via a broken delta VT scheme. Some relays are also capable of supporting the protection function via standard 3 phase VTs rather than a broken delta scheme, but that is relay dependent.

In short, I don't know of a way to do it without using some VT arrangement, as waross states, a single ground on a delta system won't change the current appreciably.

EDMS Australia

 

[OffTap100]

I for got to mention a bank of stepped resistors to both protect the bank and increase the "load" enough for tripping

 

[che12345]

Dear Mr.CuriousElectron

Q. "... with a smart relay, can a single-phase ground fault be detected ... without utilizing the typical ground fault scheme utilizing residual/zero-sequence CTs? ...It sounds like the relay can be programmed to calculate this single ground fault on ungrounded system without elaborate zero-sequence CT/VT schemes.
A1. Consider a) an ungrounded (star) [without Neutral or the Neutral is isolated from ground] or b) an ungrounded (delta) system].
When a single ground fault [i.e. dead short between one line to the round] occurs, the ground-fault current mainly (capacitive current) would be [extremely] low. BTW. this [capacitive current] exists even on a healthy system. It would be difficult for the residual/zero-sequence CTs to detect this [extremely low capacitive current]. The system maintains operational. No tripping at the up-stream or the down-dream breaker at the fault location. This is widely employed in the marine, on board of ships.
A2. A widely used detection method is having VTs where the secondary is open delta with a resistor across it. A voltage relays is connected across the resistor to detect the ground-fault.
A3. In marine applications with ungrounded/without neutral system, "earth test lamp" and "insulation resistance monitor" are used. These are indication/alarm devices. Its purpose is to alert the crew that there is a ground fault that needs attention urgently! before the occurrence of a second ground fault; which would result to a line-to-line short. It is NOT necessary to trip out the system at the first ground fault as explained above.
Che Kuan Yau (Singapore)

 

[che12345]

Dear Mr.CuriousElectron

Annexation/clarification marked in italic to my submission on 11 Jun 20 23:59
"....Its purpose is to alert the crew that there is a ground fault [in one of the phases e.g. phase A] that needs attention urgently! before the occurrence of a second ground fault [in any of the other two phases e.g. phase B or C]; which would result to a line-to-line or (phase-to-phase) short. It is NOT necessary to trip out the system at the first ground fault if the second ground fault [occurs in another location, but of the same phase e.g. phase A], as explained above."
Che Kuan Yau (Singapore)

 

[OffTap100]

I apologize. I thought we were discussing ground fault "protection". Not monitoring.

OT100

 

[CuriousElectron]

Would that have to be the first question asked early in the design process, do we want to keep operating on the first occurance of ground fault on the delta system, or do we want to trip the breaker? I'd imagine if the service continuity is important and the system is designed to withstand elevated voltagtes, only alarm would be required in this scenario..

 

[waross]

Bad idea.
Tripping the breaker eliminates the advantage of the delta system.
Running an ungrounded system, wye or delta, leaves you open to the possibility of equipment damage from superimposed high frequency, high voltages due to an arcing ground fault.
Consider the following scheme for delta ground fault protection:
Use high impedance grounding to limit the ground fault current to as low as possible.
This may be 5 Amps on a small system, 10 Amps on larger systems.
For 5 Amps on a 480 Volt system use three 3 KVA transformers, 480/?? Volts with the primaries in wye and the secondaries in delta. The delta voltage is unimportant.
This will limit arcing damage to equipment in the event of an arcing ground fault.
A "Thumper" circuit is often added.
The thumper uses a time and connects and disconnects additional resistance in the circuit every few seconds.
The ground current jumps between 5 Amps and 10 Amps.
This current is reflected in the faulted phase.
Ground faults are quickly and easily located with spot checks with a clamp on Ammeter.
The transformers are typically sized to carry between 5 and 10 Amps.
The transformers may be overloaded while the thumper is in use.

the system is designed to withstand elevated voltages

Our code requires the installation to be rated for line to line voltage.
An arcing ground fault may generate several thousand volts.
Motor windings and transformers present a high impedance to this high frequency, high voltage.
There are accounts of many motors failing all over an industrial plant when subjected to the possible effects of an arcing ground fault.
Trouble shooting may often be accomplished without an outage.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[prc]

Resistor grounding will be feasible only with star connection. In case of delta circuit( tertiary of large auto-transformers)tripping On LG fault may be ideal.I was told in this forum earlier that today’s digital relays can detect zero sequence voltage and thereby detect LG fault.Then single phase VTs with broken delta Secondaries may not be required.

 

[OffTap100]

Waross,
My apologize. I thought we were discussing ground fault detection on medium voltage systems, along with of course bank protection

OT100