PVD High Impedance Bus Diff Misop

[rogerj1]

We had a PVD High Impedance Bus Diff relay misop for an out of zone LG fault. Quite frankly, I'm not that knowledgeable about High Z style relaying schemes. This was installed years ago and we now use percentage current diff's on most all 87B schemes now. And of course its electromech and we have no other means of fault recording so there's no records to review. So far, we have verified the following:
1. All CTR's used for this relay are on the same ratio, 240:1
2. It was found there is no ground at all on the secondary current circuits . All CT's are wired in a WYE fashion, and they're all landed together in parallel for all phases/nuetrals wired into the relay.

It is my suspicion that the lack of a single ground point on the secondary circuits may have been the culprit. I've found the issue with grounding is more commonly the unexpected extra grounds to be a cause of trouble, but in this case we have no ground at all. Furthermore, I can't find any documented cases of trouble where the lack of a single common ground has caused trouble as well. Considering I'm not so familiar with the High Impedance 87B method makes it all the more confusing. I would appreciate your comments/suggestions on the topic here and anything else you could offer on the topic of missing grounds in a WYE connected secondary CT circuit.

Thanks

 

[isquaredr]

Hi
The lack of a ground hasnt caused your mal op but is a safety issue, check for the ground connection in the relay panel it may be there, but as you correctly say multiple grounds are an issue. Does the scheme have ct supervision and check zone / frame leakage, if no supervision scheme is installed start by carrying out mag curves on all ct's as a defective ct may cause a trip. If there is no check zone or frame leakage at the time of the trip were any circuits in feeder earth, stray primary earth currrent could also cause a trip if the scheme is basic. I would recommission the scheme by primary injection between units to prove stability (easy if ct test winding is wired), re calculate your setting and note your primary op values to prove setting calculations.
Good luck

 

[ScottyUK]

Check the neutral CT polarity. A reversed neutral CT won't affect stability unless there is neutral current, which in a three-wire system only occurs during a fault.

 

[norcalnewb]

How high were the fault levels? One of the issues we have had at our facility is increased available fault current has made some of the older stations buss diff settings/CT ratings susceptible to CT saturation, much like what isquaredr is saying.

 

[rogerj1]

Thanks for the discussion so far. To answer a few of the questions:
1. The relay is a 12PVD11C11A with the 87L unit at 180 Volts, the 87H unit at 2 amps. I hope that answers the current supervision question, again just unfamiliar with the scheme and we don't have alot of documentation since its such an older scheme. All CTR's at 240:1
2. Fault current contribution to the external fault (at LG fault on a line beyond the next bus) was 2931 amps as recorded by a SEL351 used for BF & 79 functions. Our short circuit model agrees with this contribution given where the fault was, but of course, this is not the actual CT circuit that the PVD relay is wired from. Those are separate dedicated CT's in all of the breakers, but they are all bushing type CT's in each of the breakers. This fault current level is certainly not a high level of concern. I have other events triggered with up to 13kA for close-in faults.

Can anyone find a manual or reference to the relay model I mentioned? The only manual I've found so far is this

http://www.gedigitalenergy.com/multilin/catalog/pvd.htm

which appears to be similar, but not the exact model we have.

For my own knowledge, what are the notable issues with the lack of a ground on the secondary circuits (other than the obvious safety reasons). How would it not attribute to faulty operations? In general, wouldn't an ungrounded wye circuit "float" and have no ground reference? I would assume LG fault conditions would be out to lunch.. ?

At this point, I'm thinking a full recomissioning procedure is needed to see if we may have a bad CT in the 87B circuit.

Thanks!

 

[cranky108]

You should also look for any open CT circuits.

You did not say which target you had, but a 87H target could indicate a bad MOV. That's the purpose of the over current unit, to be sure the relay operates for a bad MOV, rather than a non-operate.

Have you also been through the calculations for the pickup? Often it is a good idea to review relay settings, as the system can change over the years and make the settings obsolete.

 

[oldfieldguy]

You may have a CT with a lower saturation voltage then the others. I've run into this from two different angles.

In the first, the ratios, 240:1 (1200:5 here in the US) was correct for all the CT's but one set was bar-type, the others were window type. The bar-type had a much lower saturation voltage and on a through fault, they did not contribute enough current to balance the system. Result was a misoperation.

In the second case, one set of CT's was located inside the tank of an oil circuit breaker. Over a period of years, the plate on which the CT was mounted had shifted, creating, in effect, a shorted turn through the CT window. Again, the saturation was MUCH lower, <200 volts, as compared with the rest of the system which saturated at >800 volts. Again, through fault caused a misoperation.

In the first case, the CT's had to be changed out. In the second case, the OCB was opened up and the support plate shifted back to the original position.

Incidentally, the second case situation was determined in routine maintenance testing. The utility company had investigated the misoperation and never determined the cause.

old field guy

 

[stevenal]

A high impedance relay measures the voltage at the input. If the CT secondary isn't grounded, it can float up to primary potential. The input of the relay will not see this floating voltage until the dielectric strength of the relay or secondary circuit is exceeded. If a dielectric breakdown occurs at the right place, the relay might operate. Suggest grounding. I also recommend reviewing IEEE C37.110.

I had a similar experience to old guy, where an arrester bracket created a shorted tertiary and caused a high impedance bus relay to operate for a through fault.

 

[rogerj1]

UPDATE: We have found that the PVD relay was not the cause of trouble at all. Turns out, the relay never targeted or issued a trip. Initial indications from our CC classified the operation as a bus differential, but that just wasn't the case. Turns out, a cap bank relay on the same bus issued a trip during the external disturbance. However, the cap bank was denergized (the cap breaker open). The relay saw about 30 amps in the neutral of the cap secondary CT circuit. The relay issued a trip, but the unlatch trip condition was never satisfied, thus satisfying the breaker failure condition. BF from the cap bank relay triped a bus lockout. We found the secondary CT wiring for the neutral protection was grounded in two locations, thus providing an induced current condition. At steady state, the circuit was producing approx. 1 amp. We removed the ground at the breaker and the standing current goes away.

thanks to all for the discussion and help!