Step distance relaying under double circuit faults 4

[Mbrooke]

Are their any relaying concerns or set backs in regards to the simultaneous faulting of two circuits sharing the same tower? Or with other types of relaying?

 

[submonkey]

Hi Mbrooke,

Modelling mutual coupling is not that
hard, and it can be done by hand - it
is just rather tedious to check all the
different cases.

The mutual impedance has units of ohms,
or "volts per amp". In the case of zm0,
the mutual impedance tells you how many
zero sequence volts induced in feeder 1
per amp on feeder 2.

You could model this by inserting an appropriate transformer in your
equivalent circuit model.

From what I've read and seen, power system
engineers from the fifties, sixties and
seventies seemed to have a much better
grasp of fundamental theory than the current
generation.

Having said that, one could probably get
away with a lot by making (conservative)
educated guesses. Mutual coupling always
makes the zone 1 reach setting shorter and
the zone 2 reach setting longer.

Alan

 

[bacon4life]

Before widespread event recording, there wasn't much to go on for determining correct relay versus incorrect relay operation. Once we started analyzing high speed recordings, there were a number of events where the correct breakers opened, but some of the internal relay elements did not actually operate as intended.

There are also temporary faults for which no evidence was observable when patrolling the whole line from the ground after a reclose. Now, with relay distance-to-fault information, a more detailed inspection of just a few of towers may result in actually identifying very minor damage that gives good evidence for what kind of fault occurred.

 

[Mbrooke]

Hi Mbrooke,

Modelling mutual coupling is not that
hard, and it can be done by hand - it
is just rather tedious to check all the
different cases.

The mutual impedance has units of ohms,
or "volts per amp". In the case of zm0,
the mutual impedance tells you how many
zero sequence volts induced in feeder 1
per amp on feeder 2.

You could model this by inserting an appropriate transformer in your
equivalent circuit model.

Thank you, this is a good info. wye-wye equivalent transformer?

From what I've read and seen, power system
engineers from the fifties, sixties and
seventies seemed to have a much better
grasp of fundamental theory than the current
generation.

Oh yes! Not bashing anyone here, but computer programs have bypassed all reasoning, critical thinking and theoretical understanding. I myself have long been guilty of that (and still am) relying on an answer where I know nothing about how it came to be. I am blown away by what engineers were able to do 60 years ago with next to nothing in computing or the "dumb" equipment at hand.

I know after 2003 NERC became interested in load encroachment with MHO circles, but from what I have read and concluded was that in the 1965 NE back out handled the gradual voltage collapse and frequency decline far better than the system did in 2003.

Having said that, one could probably get
away with a lot by making (conservative)
educated guesses. Mutual coupling always
makes the zone 1 reach setting shorter and
the zone 2 reach setting longer.

Alan

How much guessing and on average by how much is zone 1 zone 2 being effected?

 

[Mbrooke]

Before widespread event recording, there wasn't much to go on for determining correct relay versus incorrect relay operation. Once we started analyzing high speed recordings, there were a number of events where the correct breakers opened, but some of the internal relay elements did not actually operate as intended.

And I would argue that despite that, power systems in North America and Western Europe performed very well. Theoretically in a system with generous reserve capacity, that risk can be argued as mitigated. But regardless it is worth knowing.

 

[submonkey]

Hi Mbrooke,

It is usually zero sequence mutual coupling
which has an impact. I think it would be
easiest to insert a transformer (or coupled
inductors) within the zero sequence equivalent
circuits. Sequence networks are always balanced, and so are considered "per phase".
The concept of wye wye does not really apply.

Regarding the quantitative impact of mutual
coupling - I won't give any numbers here for
fear that someone may actually use them. If you do a few setting studies for mutually
coupled lines, you will get a feel for the
typical values.

Also worth mentioning are cases where fault
current flows "forwards" on one side of a tower line, and "backwards" on the partner
line. In these cases the coupling effect is
much stronger than when one line is earthed.
It's important to understand the physical
arrangement of the lines, and where the fault
current flows. I usually look at our GIS maps
and sketch up the arrangement before commencing a study.

Thanks,
Alan

 

[davidbeach]

Under sufficiently perverse conditions you can get results such as I found several years ago. Is was setting the ground distance elements at 'A' for parallel lines to 'B'. The lines were in the 20 mile ballpark. About midway between was 'C'; and from 'C' to 'B' there were two more lines.

At 'A' the apparent impedance for a 1LG fault at 'C' was lower than for the same fault at 'B'. Hard enough to set the relays with a good model, without would have been much more difficult.

 

[Mbrooke]

Hi Mbrooke,

It is usually zero sequence mutual coupling
which has an impact. I think it would be
easiest to insert a transformer (or coupled
inductors) within the zero sequence equivalent
circuits. Sequence networks are always balanced, and so are considered "per phase".
The concept of wye wye does not really apply.

Ok, that makes more sense now, as an inductor. Transformer through me off, which would be poor in any case imho.

Regarding the quantitative impact of mutual
coupling - I won't give any numbers here for
fear that someone may actually use them. If you do a few setting studies for mutually
coupled lines, you will get a feel for the
typical values.

Post them but with a disclaimer if you feel comfortable. As far as I am concerned if someone else uses them without second thought it would be at their own risk/responsibility.

Also worth mentioning are cases where fault
current flows "forwards" on one side of a tower line, and "backwards" on the partner
line. In these cases the coupling effect is
much stronger than when one line is earthed.
It's important to understand the physical
arrangement of the lines, and where the fault
current flows. I usually look at our GIS maps
and sketch up the arrangement before commencing a study.

Thanks,
Alan

But from everything I gather all mutual coupling does is sway the single line equivalent impedance measurements somewhat rather than causing cataclysmic (for lack of better terms) under/over reach. Personally, to me my biggest concern appears to be sudden power flow reversals on the opposite line and closing into a 3 phase fault in front of the line.

 

[davidbeach]

The reversal shouldn't affect anything if the protection is just stepped distance. Start with two lines from 'A' to 'B', resulting in relays at A1, A2, B1, B2. Assume three zones of distance protection, with reaches of 85% of the line, 150% of the line, and 100%+ of the end of the next line out. Delays of 0, 20, and 60 cycles.

Now, place a fault just in front of A1. Relay at A1 will see zone 1 and trip instantaneously. Relay as A2 will see a reverse fault and do nothing. Relays at B1 and B2 will both see a zone 2 fault and begin timing.

Three to five cycles later the breaker(s) at A1 will clear. B1 will continue timing toward a zone 2 trip. A2 will now see a forward, zone 3 fault and begin timing. B2 will now see a reverse fault and stop timing.

At 20 cycles, B1 will trip and the breaker(s) will clear 3-5 cycles later. As the fault clears, B2 will still be seeing a reverse fault and still be doing nothing while A2 is about 15-18 cycles into the 60 cycle delay and never trips.

The reversal only becomes a problem with a directional based communication aided tripping scheme such as a POTT scheme. In the example above, if there is a POTT scheme the relays at B2 will be keying permission. When A1 clears the fault, there will be some period of time in which the relays at A2 begin to see the fault as forward rather than reverse while still receiving the permissive signal from B2. If A2 relays will key permissive after the reversal (for us much more likely for a 1LG fault than for a 3LG fault) then A2 could trip based on that stale receive from B2 and its own key.

In the relays we use, zone 3 is inherently reverse, so the third forward zone is actually zone 4. The POTT logic in the relay includes a block on zone 3 and a drop out timer to hold onto that block for some amount of time (a few cycles) to allow the key from the remote end to drop out.

 

[Mbrooke]

Would the breaker clearing be 2 to 3 cycles? I know nit picking, but just want to see what you have in mind.

 

[davidbeach]

A cycle or so for the relay and 2-3 cycles for breaker time plus, maybe, a bit of slop. I like to list the longer end of the time range but I do see the occasional fault that is cleared in 3 cycles (relay + breakers) with a 3 cycle breaker. But I also see a 3 cycle breaker sometimes take an extra cycle or two if it's been closed for a very long time and the go just 3 cycles on the second trip.

 

[piterpol]

MBrooke, you can find among SEL Application Guides that identified with number AG95-29, Dated on 20140611, and authored by Armando Guzman, Jeff Roberts, and Karl Zimmerman, Volume I, entitled "Applying the SEL-321 Relay to Permissive Overreaching Transfer Trip (POTT) Schemes. Read pages from 8 to 12 and you will find very clearly explained Current Reversals under "POTT Scheme Complications and Solutions.

 

[Mbrooke]

A cycle or so for the relay and 2-3 cycles for breaker time plus, maybe, a bit of slop. I like to list the longer end of the time range but I do see the occasional fault that is cleared in 3 cycles (relay + breakers) with a 3 cycle breaker. But I also see a 3 cycle breaker sometimes take an extra cycle or two if it's been closed for a very long time and the go just 3 cycles on the second trip.

Thank you yet again, I had not considered slower tripping time on breakers that have been in service for some time. I always went buy the manufacture's stated trip/clearing times, but never thought about normal deviations during the in service life of the breaker.

The relays I will be using here have those polarity dependent high speed tripping contacts. Would you say they make any dent in the relay operating time itself?

MBrooke, you can find among SEL Application Guides that identified with number AG95-29, Dated on 20140611, and authored by Armando Guzman, Jeff Roberts, and Karl Zimmerman, Volume I, entitled "Applying the SEL-321 Relay to Permissive Overreaching Transfer Trip (POTT) Schemes. Read pages from 8 to 12 and you will find very clearly explained Current Reversals under "POTT Scheme Complications and Solutions.

I found the paper. I am reading it right now and its exactly what I've been looking. Thank you

 

[Mbrooke]

One final question. When both circuits fault, one is contributing to voltage drop which the relay is not accounting for- ie it is only aware of the drop caused by its own measured current (faulted circuit). Does this under any scenarios or phase fault combinations cause under/over reach?

 

[Mbrooke]

Do any of you guys lower your K0 factors to take into account when the neighboring circuit is out of service and grounded down for repair?

 

[bacon4life]

Nope. Should we be? Does the type of grounding make a difference? In some cases, there is a single ground at the worksite. In other cases, grounding switches are used resulting in two or three grounds on the line.

In other areas of protection, we have been moving towards standardized settings to reduce human error. Creating custom settings or using multiple settings groups adds points of possible failure, so they are only used where there is a clear justification.

 

[Mbrooke]

Theoretically there may be a need since the mutual coupling properties change when the other other circuit is grounded down.

 

[stevenal]

I think the concern is where multiple grounds are used causing a zero sequence path.

 

[Mbrooke]

Line grounded at both ends for servicing.

 

[stevenal]

Suggest using single point grounding at the work site. Grounding the ends effectively energizes the parts of the line away from the grounds, and does nothing to enhance safety.

 

[davidbeach]

Under the right conditions, an out of service line grounded at both ends can see thousands of amps coupled into the line from adjacent lines in the right of way. On a 180 mile 500kV line, the middle line of three in the right of way, I modeled a fault at the bus at one end with both ends of the line grounded. With one of the ends faulted we get nearly 700A per phase, nearly 2100A of 3I0. With the other end faulted, the phase currents go to nearly 800A, nearly 2400A of 3I0. I don't have a sufficiently detailed model to calculate the standing currents that would exist simply because the lines aren't transposed.

Ground it to get the trapped charges off the line to facilitate placing the working grounds around the work zone and then unground the ends.