xfmr diff relays 2

[ventuno]

I'm not an engineer so bear with me on this question.
We have 3 xfmrs in a Substation; 2 are 10,000Kva and 1 is 5000Kva. 13.8 Kv Primary/ 4160 v secondary. They can be tied together via bus tie breakers on the 4160 side. We changed the CT wiring at the 5000kva and used the Line Diff ct's for the Xfmr Diffs and vis a versa. We moved the wires from the terminal block using corresponding numbers,(3C1 to 1C1, 3C2 to 1C2, 3C3 to 1C3, 3C0 to 1C0).
We energized the transformer primary first and as soon as we picked up the secondary main we had an 86 Lockout from the 87T relays(HU). Since we had noticed a wiring problem on the terminal block we tried moving the wires in different combinations(123,312,etc.) until the breaker stayed in.(I know this is not how to do it but the original installation had 4 problems in the xfmr diff scheme and we had no prints or wiring diagrams.)
We showed about 200amps on the secondary which was tied to the bus along with (1) 10,000kva transformer, the other was being repaired for leaks. We kept the 5000Kva on the bus for about 15 minutes with no problem.
Later the other xfmr was energized and leaked more than before, so our boss put the 5000kva on line and had diff lockout with Phase 1&3 targets. Tried again Phase 2 target and lockout.
We are going to check ct wiring tomorrow. Could we have a polarity problem due to the change from a ct in the rrear of the breaker to a ct on the load side stabs of 13,8 kv breaker.
thanks for any suggestions.
Ron

 

[Bung]

Old electro-mechanical transformer diff relays rely on the correct connection of the CTs to correct phase shift in the main transformer, and also to filter zero phase sequence currents for out of zone faults. You don't say what your power transformer vector group is.

For example, a Dyn1 transformer will have its primary side CTs connected in star and the secondary side connected in delta. The CT ratios must also be chosen to match each other taking into account the voltage ratio. The star connected CTs will need a sqrt(3) factor to account for the d-y transformation. I suggest youget somebody who knows what he is doing in to advise - it may cost a few hundred bucks, but could be worth it in the long term.


Bung
Life is non-linear...

 

[busbar]

You might look through
IEEE Standards
C57.91
C37.103
C37.110,
C57.13.1 and
C57.13.3.

 

[stevenal]

Could we have a polarity problem due to the change from a ct in the rrear of the breaker to a ct on the load side stabs of 13,8 kv breaker.

Yes, these two sets are generally set up with opposite polarities. Why the change? You usually want to overlap zones of protection across the breaker contacts.

I'm not an engineer...

Hire one maybe? Your trial and error approach doesn't sound very efficient.

 

[ventuno]

to Stevenal,
the person who created the problem claims to be an "engineer" and I'm just a Relay Tech trying to make the best of a bad situation. He doesn't want to pay for a real engineer to come in and solve the problem he wants to replace the HU's with a Multilin relay that can be programmed to accept the wiring as found. Is this true?
Anyway I'm on vacation starting today and it will be waiting for me when I return. Thanks for the info so far.

 

[Bung]

It sounds like nobody has any idea what the wiring is now, so what are you going to tell the relay? The Multilin relay (assuming its a SR745) can do the vector and ratio correction, but only if it is told the right thing. It ain't magic. I would suggest that somebody needs to do a full CT wiring check to find out exactly how they are connected. Then the rest is easy, you just program the SR745 to suit.

Doing it by guess and by God is likely to get somebody hurt. If you fancy going home in one piece, get somebody competent to assist, or at least refuse to touch it.


Bung
Life is non-linear...

 

[busbar]

At this point a phase-angle meter would be worth it’s weight in gold, even for a Multilin ‘do everything’ box, with all portable-instrument measurements plotted on paper—referenced to Van or Vab.

Barring that, however rude it may seem, “close and test” after phase rolling can be a risky proposition. Be sure that you understand exactly what the senior “engineer” wants—on a fundamental, step-by-step basis, explained by him to you in a “talk-to-me-like-I’m-four-years-old” manner. Then, as stupid as this sounds, repeat each instruction back to him, confirming as much detail as possible. If this guy insists on being at the helm, make darn sure he unerringly realizes that before proceeding any further.

 

[dpc]

Well, it seems likely the CT wiring is incorrect, but the challenge is in figuring out what is wrong.

There must be about 500 ways to hook up six CTs to a transformer differential relay and only one of these ways will work. With the old HU relays, you are either have to ring out each CT lead from one end to the other and check each CT polarity or energize the transformer with the relay out of service and check CT currents with phase angle meter or scope.

With a Multilin SR-745 or other modern digital transformer relay, you can disable the differential trip function (leaving overcurrent functions in service), energize the transformer and check the current phase angle using the relay software. In many cases you can display a phasor diagram of the six currents. Because of the phase shift in a delta-wye transformer, you will have to consult the relay manual to see what a proper phasor orientation should be, but it's much quicker than just swapping wires at random.

It's better to get it right the first time, but the digital differential relay makes troubleshooting a heck of a lot easier. Just don't forget to re-enable the differential function when you're done.

 

[ventuno]

I have checked the wiring of the CT's in both breaker cubicles and will be able to verify polarities next week. We just found a set of Prints for the circuit in question, they were in a file with a different name than the actual name we use for this Substation. I will be able to check my sketches of the wiring with the instruction manual for the relay and the wiring diagrams that we now have.
I just noticed re-reading Stevenal's Post that we may now have a small gap from the incoming line to the load side of the breaker that has no differential protection. Is this true?
Is this a big problem? We have three 50/51's; one 50/51N; an 87L pilot wire and the three xfmr diffs on the primary side breaker cubicle.
Thanks to all for the good info. I will proceed with caution and from your posts I will verify polarities and try to determine Phase angles if possible.
I'll post again in about 10-12 days after I get back to work and have a chance to go over the circuit armed with prints and your input.

 

[ScottyUK]

From the sound of this installation, and the amount of intrusive work that has obviously been carried out, a primary injection test would be very prudent. You will probably need to hire a specialist.



------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[advidana]

Some times the engineer in charge blow it. Maybe his boss will hire a better engineer after this costly mess. Good luck on your vacation.

 

[digitrex]

Better follow the advice from:

dpc: only one way of CT connection will work.
ScottyUK: Do a primary injection test to verify diff protection scheme.

My suggestion: (How do you do a priamry injectino test from primary to secondary windings in order to include CT of both side?)
Connect a 3-phase LV test supply(no joke, a normal LV supply 200V or 400V) to the HV side of winding(NOT LV side, otherwise the test cable will get overheated due to too high current or it will electrocute you at the HV side) while the LV side is solidly shorted together. Primary current will not be very high under this test but it is enough for you to check whether the diff protection relay is detecting unbalance current.