Diesel generator differential protection trip help 1

[scott88]

Hi,

We have a problem with our emergency diesel generator. The generator is a 1100kVA 400V 50 hz machine, manufacturer Leroy Somer with Siemens SIPROTEC 7UM62 relay protecton. On running the generator up it is stable and running fine for anything onload between 40minutes to an hour and then trips on differential protection. The generator has CTs on each phase situated just before the circuit breaker and before the star point. We have tested the generator windings and all are balanced and low, mega checked to earth and between phases all with readings in the Gohms, the CTs have been checked and tested and are also fine.

The readings are.
SIDE 1 (just before the circuit breaker)
L1 - 1.01kA
L2 - 1.02kA
L3 - 1.02kA

SIDE 2 (Just before the star point)
L1 - 0.72kA
L2 - 1.02kA
L3 - 1.02kA

Does anybody have any ideas or advise for further tests or fault interogation?

All other protecion functions are fine and show know pickups or cause for concern.

Thanks,
Scott

 

[waross]

Is the problem in the windings, the CT, the CT wiring or the relay?
The windings and the CT seem good. Next check the wiring for high resistance.
Consider rolling two phases and see where the problem goes. If the issue "follows the roll" the problem is on the machine side of the point at which you rolled the connections. If the issue on the same phase, the problem is on the relay side of the "roll".

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

 

[ZapSib]

Scott,
Is there anythink happening in th esirrounding network when 87T trip ?
I mean, is there any Tfo or Capacitor bank switched in ?

 

[electricpete]

Obviously, the possibilities include: generator fault, ct fault, ct secondary circuit fault.

If it was a generator fault ct's/relays working fine, then you had 300A flowing to ground, but if left no evidence detectable by ground or megger check of the machine. I consider this unlikely, therefore I'd focus more on the other two (ct fault, ct secondary circuit fault)

One check of ct circuit: For a system with electromechanical relays, I would lift the ground at the wye point of the ct circuit (relay end), and megger the ct circuit. For solid state /electronic relays, you also have to isolate those devices or ensure they will not be damaged by voltage applied. Replace lifted ground.

Lift the leads at the ct's, check resistance of each path and see if the same (may not be a good test for electronic relays).

Perform excitation test and ratio test of the CT's.

Check the relays.

With machine running, many relay engineers like to check current magnitudes and angles.



=====================================
(2B)+(2B)' ?

 

[electricpete]

current magnitudes and angles.

That would be ct secondary current.


=====================================
(2B)+(2B)' ?

 

[electricpete]

Another thing to look at is whether the current combined with ct secondary impedance can be pushing your ct's into saturation.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Although true fault is unlikely imo, for a large horizontal machine, it is typically not hard to pull panels (just above bearings) in the end covers to inspect the inventions for obvious visual damage. Also inspect other easily accessible areas like terminal box.

=====================================
(2B)+(2B)' ?

 

[scott88]

Hi,

Thanks for the responses. I will keep you updated of any findings. We have performed terminal box inspections and general inspections. A differential of that size one would expect to find a fairly visual fault if it was the machine. We will change the phase of L1 CT today and see if the fault follows this to conifirm if it is relay side or actual machine fault.

Thanks

 

[rasevskii]

@scott88:

You might want to consider running a primary test: that is a short circuit at the end of the farthest CT set. But you would have to disable the AVR and run the unit on separate excitation to the exciter field from a variable DC supply. If the diff protection has a reversed CT, phasing wrong or a programing mistake in the SIPROTEC, this will find it out.

You would need a SC current of around 40 to 60% of rated to find the fault.

Likely such a test was never done at the original commissioning.

rasevskii

 

[electricpete]

Have you ever gotten the machine to any load (current) above 1.02kA without tripping in the past?

How high have you gotten it?

=====================================
(2B)+(2B)' ?

 

[waross]

Another check is the current in the grounding connection from the wye point to ground. No current indicates a measurement fault. That is CT, wiring or relay.
If you find the missing current your stator winding may be failing to ground in the last turn or so. Remember, the purpose of differential protection is to detect such faults.
I have seen intermittent faults on generators which only occurred under working conditions but not under test conditions, even when heated to operating temperature in an oven.

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

 

[rasevskii]

It could be a faulty VT that fails to ground after heating up under voltage for a while.

rasevskii