Differential Protection Coordination 1

[nickoliver]

I am dealing with a nuisance tripping problem on a differential protection scheme and am looking for some help in understanding as this is a bit over my head. (We do have an engineer working on the problem as well)

The situation is we have a line up of GE 15KV switchgear with A and B side feeds and an automatic throw over system which among other things feeds 6 substation transformers (12.47KV to 480/277). A and B systems are redundant with each side having one 3000KVA, one 2500KVA and one 1500KVA transformer. Each breaker has differential protection with the primary CT’s located around the load side bus of each 12KV breaker and the secondary CT’s located around the bus stabs of the breakers in the 480V gear after the transformers. Each 12KV breaker has a T60 relay providing the protection. We ran interconnecting CT wiring from the secondary CT’s to the T60 relays. The problem is the 2 breakers feeding the 1500KVA transformers keep tripping on current, percent differential when only about 175A (@480V) of load is applied. One of the 2500KVA transformers is having the same issue.
What we discovered through some investigation is that the three breakers that are not nuisance tripping have CT ratio’s configured as follows:
2 3000KVA transformers have primary CT’s at 300:1 and secondary CT’s at 4000:1
1 2500KVA transformer has primary CT’s at 300:1 and secondary CT’s at 3200:1
The breakers that are tripping are configured as follows:
2 1500KVA transformers have primary CT’s at 100:1 and secondary CT’s at 2000:1
1 2500KVA transformer has primary CT’s at 100:1 and secondary CT’s at 3200:1.
It looks as though changing the 100:1 CT’s to 300:1 may fix our problem. Does this make any sense??
Thanks in advance for any input.

 

[stevenal]

The actual CT ratios must match those programmed into the relays. If they don't match, one or the other must be adjusted. The T60 should be able to provide all the information you need and more to correct this. Wrong phase shift compensation, polarity, crossed phases, shorting screws left in, etc can all cause nuisance tripping.

 

[dpc]

No, it really doesn't make any sense. The CT ratios are what they are. As stevenal says, the CT ratios programmed into the relay must match the actual CT ratios.

Your symptoms indicate a problem with CT polarity and/or phase in the circuit that are tripping. Due to the restraint in the differential relay, the tripping will not occur until there is some load on the transformer. If you observe the current in the relay at low loads before it trips, you will be able to see the differential current, and you can probably get enough information to draw the phasor diagram that will show the problem.

The T60 can provide you with the cause of the trips as well as the current. If it is tripping on differential, the relay will also indicate what it thinks the differential current is.

This should help point you in the right direction. There are about 30 ways to hook up CTs into a transformer differential relay, but only one is correct.

 

[JensenDrive]

What is the full ratio of the CTs? Are you tapping them way down? You might be amazed at how high the excitation current can be when you tap CTs way down. I have seen it cause xfmr diff trips.

For example, the knee point excitation current on a 1200:5 CT might be 0.2A, but the knee point excitation current at a 100:5 tap is roughly 12x as much, or about 2.5A, and is no longer neglible. Furthermore, when you are on low taps, the knee point voltage drops (e.g. by 1/12 in this example) so it is much easier to reach the point where the CT is not a very good CT.

 

[nickoliver]

The CT ratios match the ones programed in the relays. We confirmed that last Friday. Though we did find that the 2500KVA transformer was set at 300:5 in the relay (where it should be) and the CT's are tapped at 100:5. The relay was changed to mach the tap at 100:5 which is different than it's A bus counterpart. I confirmed today it still trips though on a slightly higher level. I suspect tomorrow we will be changing the tap and relay to 300:5.
We have physically checked polarities and wiring more than once and are confident that that is ruled out, unless of course our connection diagrams are incorrect. The T60 will produce a phasor diagram but it will only do it when there is current flowing. We have had a hard time working with mechanical to get the right number of pumps running so we can see and record it and the relay tripping. Hopefully tomorrow. We also had some software issues we are trying to work out with out IT department as far as the required software being installed.
I don't know the full ratio of the CT's. I will have to check tomorrow but I know 100:5 is at the low end of the scale. That is an interesting point since the 300:5 breakers work fine so far. We might find they still trip just at a higher level.

 

[apowerengr]

Sounds like you need a testing and commissioning agent who understands differential schemes - they'll quickly tell you where the problem(s) are and will be well worth any fees they might charge. Was this system installed and commissioned internally or through a third party?

 

[oldfieldguy]

I am going to agree heartily with "apowerengr".

It sounds like a fairly straightforward CT polarity issue. Many filed people lose track of the issue of CT polarity in differential circuits.

I have seen the drawings wrong, and I have seen the wiring wrong, and I have seen the CT's installed backward from the drawing, and I have seen the wiring going to the wrong set of CT's. I have also seen the incorrect settings in the relay.

The T60 is a wonderful platform, but some people get confused by the relationship of current inputs and definition of sources in the setup of the relay.

I have also seen problems where the relay logic called for trip on the status of the differential element pickup as opposed to the differential element operation. There's a difference between the two. I made good money finding and explaining the difference to a client and his contract engineer.

I'd love to help you figure this thing out. All it takes is time and money.

old field guy

 

[jghrist]

You should be able to see any CT wiring errors or polarity problems with the phasors at loads below the tripping point.

 

[oldfieldguy]

jghrist is right!

In troubleshooting some of these problems, false tripping when we KNOW the transformer is good, I have disabled the tripping function of the differential element for the time you're testing.

On the T-60, this is easily done in the programmable logic. Also, with the T-60, you can route the output, again in the programmable logic, to one of the user-programmable LED's to see when the offending element operates, and you can snag the event record for analysis when it operates.

Then you can load the system up in the configurations that are causing problems and use the relay's own data in the vector analysis functions to see what changes and when, to cause the trip.

Like I said, this is fun stuff!

old field guy

 

[nickoliver]

I have a little more info. We changed the tap and programing on the 2500KVA transformer to 300:5 and were able to bring the load (480 side) up to about 350A with no trips. We were able to look at the vector diagrams in the T60 software and an interesting thing was happening. The primary and secondary vectors were perfectly arranged but the primary vector keeps disappearing. It comes back for an instant then disappears again.
One theory that came up is the fact about 98% of the load is VFD's. Someone had an issue with feeding VFD loads with electronic relays. Their solution was to rearrange the loads so more linear load was mixed with the VFD's. We do not have that option.

 

[RalphChristie]

"The primary and secondary vectors were perfectly arranged but the primary vector keeps disappearing."

Is this a new installation? And have you tested the differential scheme? (CTs, wiring, etc)

Possible causes, especially on the HV-side's CTs:
Saturating CT/CTs (do a mag-curve test on the CTs)
Fault in the wiring. Something like an intermittent earth fault, flash over, or even a loose connection somewhere in the circuit. (do a megger-test on the wiring and re-check all your connections. Check especially if the test-blocks is fitted securely)

Regards
Ralph

[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[tinfoil]

Further to RalphCristie's post:

You have not told us what your CT ratings are. Using the ANSI system, the CTs would have a rating like C100, C400 or something.(The european spec would have a VA rating).

The C value is the amount of voltage the CT is designed to sustain (during the duration of a fault detection) with accuracy (i.e. not saturating) across its FULL secondary, so a 1200:5 C200 CT can sustain 200V across its two outside taps (i.e. greatest ratio). You have to ensure that your secondary impedance (including secondary wiring and connections) times your worst-case secondary current does not approach this value.

The ideal formula is that required C > (1+X/R) * (Imax,sec *Zsec)

where X and R are the primary system reactance and resistance.

Note that when you choose a reduced tap, you must derate your CT as well. If you were to choose a 300:5 tap on the CT above, your effective rating on your CT becomes 300/1200* 200 = C50.

However, the 'acceptable' rule-of-thumb for modern relays is that C > 2 * Imax,sec * Zsec


Note that in nearly every real system, this is not usually an issue under load currents, just for fault level currents.

 

[dpc]

The T60 relay allows you to designate which incoming source is used as the reference signal for determining phase angle. This can be accessed under Power System settings. I'd make sure you are using a Primary source as the reference. If the primary phasor continues to disappear even if set as the reference, then I might suspect some intermittent wiring issue. You can also try, as a last resort, turning off the Frequency Tracking feature in the T60 to see if this helps. If the secondary current is highly non-linear, it's possible the relay is having difficulties locking in the fundamental secondary currents.

You need to verify that the relay is properly set for the actual transformer phase shift, and also verify that the actual phase sequence (ABC or ACB) is correctly set.

I know you are tired of hearing it, but experience tell us that a CT wiring error or relay configuration error is the most likely culprit. Guilty until proven innocent. Often multiple wiring errors can mask themselves under low load conditions. So I would double and triple check all CT polarities, phase, and wiring into the relay.

If the relay is tripping on differential it is because it "sees" a difference in current between primary and secondary. You just need to figure out why. Looking at the relay's computed difference current and plotting that on the phasor diagram can often give clue as to the problem.

 

[davidbeach]

Lift the one ground on the CT secondary circuit (or one per side of the power transformer) and verify that you don't have any other connection to ground on the CT secondary. Multiple grounds will always cause trouble.

 

[slavag]

Hi Nickoliver.
Next option of problem. CT secondary wiring size.
Pls. check T60 CT requerements and calculated n' .
Regards.
Slava

 

[nickoliver]

OK I think we got it. We had the T60 Guru from GE out today and it was a combination of many of the suggestions above. On two of the transformers, the 1500KVA, the drawings were wrong and the phase shift was wrong. We swapped some CT leads at the test block to push secondary current through the relay the opposite direction and changed the phase shift from -30 Deg to -330 Deg. The T60 shows 0 differential current under normal operation and everything works great. The other transformers had all the polarities correct but the phase shift needs to be change to the same value as above. Thanks for the suggestions.
Nick

 

[JensenDrive]

Thanks for coming back to tell the results, and getting a good conversation going.