Earth/Ground Fault Protection

[genhead]

A small (3MW), remote area power station consists of several generators at 415v (wye) stepped up to an 11kv (delta) distribution system. The 11kv grounding is via an grounding transformer and resistor (10 amp).
If an earth leakage relay was installed with its CT over the 11kv neutral conductor of the grounding transformer, would this provide adequate protection for earth fault ?
If not, what would be a satisfactory solution ?

 

[jbartos]

Suggestion: It appears that the grounding system described in the above posting resembles the high-resistance system grounding. This grounding scheme is dependent on the Insulation Level or Class of cables:
100% insulation level or class must have an alarm with trip timed up to one minute, i.e. 60 seconds. Else, the cable insulation may be damaged if the ground fault develops.
133% insulation level or class must have an alarm with trip timed up to one hour to prevent cable of any insulation damage.
173% insulation level or class must have an alarm only; however, the trip is not necessary since the fault may stay on for an indefinite period of time.

 

[peterb]

Depending on the sensitivity of the scheme, the neutral CT would generally provide adequate protection for earth faults on the 11 kv system. Are there any downstream 11 kV branch circuits? If so, then some form of discriminating earth fault protection would be required on each feeder.

As far as the alarm vs. trip situation raised by jbartos is concerned, this is dependent on whether or not the earthing resistor is continuously rated. The resistor time rating will determine the maximum trip time allowed for the scheme. In any case, the cables will need to be rated for full line-line voltage, for use on a non-effectively earthed system.

 

[busbar]

robint, for your scenario, would you describe specifically what the earth-leakage relay will control?

 

[genhead]

1. Busbar, The earth leakage relay will trip the lv generator circuit breakers. The system as it stands is very primitive, with no earth fault protection on the individual 11kv feeders. In the event of an 11kv earth fault, the entire system will be lost.

2. Is it normal practise to cable the earthing transformer directly from the 11kv bus, with no protection ?

3. The resistor is rated for 10 seconds.

 

[peterb]

1. You may want to think about installing earth fault protection on the feeders - if these are cable circuits, a zero sequence CT can be installed over the cable and connected to a definite time overcurrent relay. You will of course need sensitive protection for the low level of earth fault current.
2. Normal practice would be to fuse the earthing transformer for phase fault protection.
3. A 10 sec rated resistor suggests that you may want to consider a definite time delay trip of no more than around 5 sec for the earth fault relay, with a pickup level of around 3 - 5 A.

 

[busbar]

A fuse operation on a grounding-transformer lead may introduce damaging resonance problems. A ground fault in the 11kv/415V transformer or on the 11kV line will likely be detected by the ground-fault relay, but there’s nothing to remove the fault. If the fault is fed from the other end of the 11kV circuit, there is nothing local to disconnect the resistor. It sounds like a communications channel for transfer trip is probably miles out of the budget.

 

[DEPU]

Is there any reason why your system is the other way around?
The usual connection for a transformer connected to generator is delta on the generator side & star on the load side.

 

[Bung]

Is the grounding impedance predominantly inductive (ie due to the grounding transformer) or resisitive? You may need to make sure you don't have a reactively grounded system (X0>10*X1) to avoid nasty over-voltages on arcing earth faults. But to answer your question, yes, it should. It would just be a matter of deciding on suitable settings for the relay.

Bung
Life is non-linear...