Primary Injection Test of Earth Fault

[shkim2000]

I need to know how earth fault test for MV switchgear is carried out by injecting primary current.
The power system is 6.6kV 3phases with isloated neutral point. The customer has doubt that breaker is tripped by multifunction relay flowed vi CT and thus actual earth test is recommended. How do I prepare for earth test such as making some shortciuit, and injection point, and etc.
In addition to that, how do we set generator differential protection in line with CT's function ?

 

[Andiri]

For earth fault testing, inject across 2 phases and short them out on the P2 side of the CT's (ie 1 lead of test unit to P1 of say red, the other to P1 of say white, then connect the P2's of red and white together with a suitably sized lead). Avoid using flexible aluminium conductors/leads for shorting, rather copper; found Al impedance interferes with injected current.

Making the smoke come out is easy; getting it back in is a bugger.

 

[shkim2000]

I noted how earth between phase and pahse is done.
But, if I make one phase cable jumper to earth before energizing in 6.6kV acutal power network system after that I switch it on VCB to see if VCB is tripped due to earth fault, to confirm another earth fault(one phase to earth) by ZCT in isolated neutral system. What do you expect any possible hazard and damages ?
The reason is that the customer doubts that VCB trips due to earth fault on neutral floating system.
If there is any alternative to confirm it, please advise me
Please share your experience and technical advice with me ?

 

[RalphChristie]

If you just inject with a primary testset through 1 CT (or through 2 CTs like Andiri suggested) you'll find a earth-fault trip. But you have to simulate a fault keeping in mind how your system is going to behave during real fault conditions. It doesn't help to inject a current of say 2000A through the CT, but under a fault-condition there will just 200A be flowing.

You have to know your system parameters and how it reacts during a fault.

The problem with primary testsets is that they have a very low output voltage, thus how higher your resistance the more difficult it become to inject the current. (I'm talking of milli-ohms if I talk about resistances)

I doesn't really understand how you want to do your test, but if your system is floating, (neutral not earthed) there won't be any earthfaults before two phases is going down to earth. Can you be more descriptive on how your system look like? (voltages, earthing methods etc)



Regards

Ralph

 

[shkim2000]

My system is non neutral earthing system from 6.6kV power generator to distribution system. There are feeders to 6.6kV/440V transformer with ZCT. The protection system is to trip VCB when one phase is earthed or let say low insulation before two phases are earthed or short-circuited. The protection relay is to monitor and to trip current flowing ZCT.

 

[busbar]

 
Seems that compliance voltage of a primary-injection test may be high given a high "loop"-reactance component of typical staged-fault impedance.

Classically, an ungrounded system must have a “preexisting” ø-g fault, and a second on another phase producing tripping on a ø-g-ø basis.

 

[Interrupter]

We use high-resistance (10,000 Ohms) grouding on 6kV system. Grounding is made with EVT.

Upon an earth fault, fault current through ZCT is supplied from the capacitance of the system (mainly cables) plus 6.6kV/1.73/10000=0.38A. With this arrangement, ZCT trip point is set so low, 0.2A to 1A. With this low setting, relays are all directional (67G)and work with grounding overvoltage realy (59N) on the system. If OCG relays are used and set so low (0.2A), they may trip for a fault on different feeders due to cable capacitance of its own.

To test the system, you need to utilize a water-resistor and activate a real fault through the resistor in order to confirm 59N+67G operation. If it is OK for client, you can use a high-pot tester and do some wiring around ZCT activating 67G, apply AC voltage to the EVT separately for 59N operation.

I would prefer minumum 100A grounding against isolated system. You would set your earth fault relay 5A to 10A, then primary injection test can be easily done by circling wire around ZCT and inject small current into the circuit. Much simpler in design and testing.

 

[digitrex]

shkim2000,

I believe you are asking whether generator's earth fault protection (of the multifunction relay) is able to function correctly when that generator's neutral is not grounded.

Is there any other neutral point in that MV system that is grounded? If there is a grounded neutral elsewhere, an earth fault on any phase will have flow of earth fault current. In this case that generator's earth fault protection would because Restricted Earth Fault Protection and it only detects generator's internal fault.

If the whole of that MV system does not have any grounded neutral at all, all current sensing earth fault protection within that MV system will not work. You can only apply a residual(zero sequence) voltage relay to detect an earth fault.

I don't recommend ungrounded system as it is prone to transient overvoltages.

 

[raisinbran]

Shkim2000,
This might not qualify as primary injection, but several of our CTs are wound with two secondary windings - one to feed the protective relays and the other is a test winding. The ratio of the transformer would be specified as 600:5:5 (for example). All of our testing is done using the test winding, which exactly simulates a true primary injection. This testing is done at low voltage with a Multi-amp or other variable AC current source.
I apologize if this post is off of the topic.
Regards,
Dave

 

[RalphChristie]

Sorry

I'm not familiar with the term ZCT. Can someone explain it to me? Am my assumption right if I say it is a type of balancing core or ring CT?

Thanks

Ralph

 

[digitrex]

shkim2000,

Your customer is trying to tell you that there will be no earth fault current flowing in an ungrounded system. When there is no earth fault current flowing, the CTs will not pick up anything. He is now challenging you to apply an actual earth fault on a live 6.6kV system to prove that an earth fault can operate the protection relay.

He is going to drive you to a corner.

 

[digitrex]

And, I agree with your customer.