Hypothetical question (line differential protection)

[Bugman1400]

Sorry, I meant LD. :ashamed1: Although I did not know PLC does not support LD. I guess I should ask what is PLC is used for?





Correct, and if using a communication (POTT) scheme once zone 1 is cleared the remote terminal can be tripped by communicating that a breaker has opened tripping instantly afterwards eliminating the time delay in none communicating schemes. However, where LD comes in is easy of setting, application, and possible faster tripping of all involved breakers. No need to do extensive impedance calculations on fault current in relation to distance and guessing what values to plug in.





Good point! That should have been obvious now that I think about it. Though to be fair to myself every breaker and half I've seen had a breaker failure scheme.







Well, that depends on the utility. Some utilities will tap directly off of 115kv lines many times, while others loop the lines in and out of each substation leaving only 2 and 3 terminal 115kv lines. In this discussion we assume that all 115 and 230 kv lines loop in and out of each transmission to distribution substation.


However, that is not to say it can not be done. Could the summed currents not take into account a inverse time current curve? SEL offer relays that can integrate line taps into the differential application. Both SEL and GE have papers on it, though I can not find the SEL at the moment.


http://store.gedigitalenergy.com/faq/Documents/L90/GER-3978.pdf






They do, often primary is LD with secondary as POTT or DCUB. But would it not be possible to have two redundant relays each with their own communication channel doing LD?

PLC (Power Line Carrier) is an old, established method for sending a Trip or Block signal over to the other side. It consists of using a carrier signal that rides on the main conductor along with the power flow. The information carried in the signal usually consists of a digital type of ON/OFF or TRIP signal. Since LD would require a much higher bandwidth of information, PLC cannot be used. PLC is being used less and less for new construction.

You are correct that the tapped loads off a line can be summed (or compensated for) but, for this to happen, comm must exist between the tapped subs back to the LD relays. This is not always practical due to economics or feasibility. However, for new construction (including tapped loads), this is more technically feasible but may not be economical. I am not aware of using overcurrents to integrate into the LD. As has been done for many years, if the tapped load is light enough, it is possible for the LD to be de-sensitized enough to allow a non-compensated tapped load. However, this has been debated by relay engineers for many years since most engineers don't like the idea of de-sensitizing an LD.
Also, I think you should think about why a 500 or 345 or 230kV application would have an LD for the primary and a POTT or DCB for the secondary when comm is available for both relays to the other side. I think it is typical for most relay engineers to avoid a thing called Common Mode Failure (CMF). In the case where you would have LD on both the primary and secondary relays, the potential exist for CMF to exist. It is always best to apply several diverse layers of protection for the best result.

 

[mbrooke]

PLC (Power Line Carrier) is an old, established method for sending a Trip or Block signal over to the other side. It consists of using a carrier signal that rides on the main conductor along with the power flow. The information carried in the signal usually consists of a digital type of ON/OFF or TRIP signal. Since LD would require a much higher bandwidth of information, PLC cannot be used. PLC is being used less and less for new construction.

Got now it and well explained Thank you.

You are correct that the tapped loads off a line can be summed (or compensated for) but, for this to happen, comm must exist between the tapped subs back to the LD relays. This is not always practical due to economics or feasibility. However, for new construction (including tapped loads), this is more technically feasible but may not be economical. I am not aware of using overcurrents to integrate into the LD.

I might be wrong, but I have heard of relays that will do just that.

As has been done for many years, if the tapped load is light enough, it is possible for the LD to be de-sensitized enough to allow a non-compensated tapped load. However, this has been debated by relay engineers for many years since most engineers don't like the idea of de-sensitizing an LD.

Would desensitization not make a different up to a certain point? My understanding is most faults involve hundreds of amps or more (high impedance faults are rare at transmission voltages) making it possible to reduce sensitivity without compromising protection?

Also, I think you should think about why a 500 or 345 or 230kV application would have an LD for the primary and a POTT or DCB for the secondary when comm is available for both relays to the other side.

Im guessing in the event communication was lost?

I think it is typical for most relay engineers to avoid a thing called Common Mode Failure (CMF). In the case where you would have LD on both the primary and secondary relays, the potential exist for CMF to exist. It is always best to apply several diverse layers of protection for the best result.

Wouldn't independent communication channels for both DF relays eliminate common mode failures?

If you mean different relays (DF relay + POTT relay), I would argue that having layered protection increases the risk of failure over DF relay + DF relay in that more setting exist and in term more chances of setting one incorrectly.

On of the beauties of DF is that its much simpler, almost plug and play when compared with other schemes.

 

[Bugman1400]

Got now it and well explained Thank you.




I might be wrong, but I have heard of relays that will do just that.






Would desensitization not make a different up to a certain point? My understanding is most faults involve hundreds of amps or more (high impedance faults are rare at transmission voltages) making it possible to reduce sensitivity without compromising protection?





Im guessing in the event communication was lost?






Wouldn't independent communication channels for both DF relays eliminate common mode failures?

If you mean different relays (DF relay + POTT relay), I would argue that having layered protection increases the risk of failure over DF relay + DF relay in that more setting exist and in term more chances of setting one incorrectly.

On of the beauties of DF is that its much simpler, almost plug and play when compared with other schemes.

You are correct that bolted faults result in higher current in areas with a strong source. However, think about weak areas with LD or n-1 or n-2 conditions. Also think about faults that have a considerable amount of impedance in them (ie arc faults or line-to-ground faults involving trees or whatever). Sensitivity is essential!

I think you missed the point about having a primary and secondary relay that have different schemes even though they both have comm over to the other side. Many times this comm is fiber or other type comm that is capable of providing enough bandwidth for two LD schemes. However, it is atypical for both schemes to be LD because of the common mode failure concern. The concern is that there may be hidden gaps in the way that a relay applies LD that may lead to a failure in protection or to protect 100% of the line. For example, there was a paper about an SEL line relay that had an algorithm that contained a deficiency that would prevent the relay from declaring a fault for a certain type of evolving fault. So, it has been accepted by many that every relay may have some type of inherent defect and that the best approach is to use different schemes and different relay manufacturers when possible.

 

[GoldDigger]

There have also been authoritative studies of complex networks like the telephone network, the internet backbone, and the power network. One strong conclusion is that the probability of a catastrophic cascading failure is greater the more uniform the network components, including software, are. Redundancy without diversity is not enough.

 

[mbrooke]

You are correct that bolted faults result in higher current in areas with a strong source. However, think about weak areas with LD or n-1 or n-2 conditions. Also think about faults that have a considerable amount of impedance in them (ie arc faults or line-to-ground faults involving trees or whatever). Sensitivity is essential!

I think you missed the point about having a primary and secondary relay that have different schemes even though they both have comm over to the other side. Many times this comm is fiber or other type comm that is capable of providing enough bandwidth for two LD schemes. However, it is atypical for both schemes to be LD because of the common mode failure concern. The concern is that there may be hidden gaps in the way that a relay applies LD that may lead to a failure in protection or to protect 100% of the line. For example, there was a paper about an SEL line relay that had an algorithm that contained a deficiency that would prevent the relay from declaring a fault for a certain type of evolving fault. So, it has been accepted by many that every relay may have some type of inherent defect and that the best approach is to use different schemes and different relay manufacturers when possible.

But have these issues not been ironed out? And since DF is relatively simple, no much can go wrong? My issue with 2 different relaying schemes from two different manufactures is a person has at least a 4x of misapplying a setting. At least that is how I see it IMHO.

Also, in regard to different schemes, dont such become harder to rely upon during N-1 or N-2 due to reduced fault current as you mentioned?

 

[mbrooke]

There have also been authoritative studies of complex networks like the telephone network, the internet backbone, and the power network. One strong conclusion is that the probability of a catastrophic cascading failure is greater the more uniform the network components, including software, are. Redundancy without diversity is not enough.

If I may introduce something rather eye opening mentioned in an Engineering forum:

I know where it comes from, and 20 years ago it may well have been appropriate. The "use different" approach introduces a strong bias toward dependability at a definite cost in security while misoperation shows that we, as an industry, really need to focusing on security. We see that the vast majority of protection system misoperations are unnecessary trips. And of those unnecessary trips, the single most common cause is setting errors.

If I'm setting two different relays I have twice as many settings to work with as I do when I'm setting two of the same relays. If my greatest risk is a bad setting that causes an unnecessary trip I've just doubled my total risk of a misoperation. When the numeric relays were new and unproved, it may have been very prudent to take an approach of "if this one doesn't trip, hopefully that one does". But that's not the problem the industry has today. I've heard told that at one point, particularly when event analysis was far more difficult, that "well, it reclosed, so that's good" was a common attitude. Now we have PRC-004 and the requirement to analyze every operation to ferret out all of the misoperations and correct them. Security failures seem to be very low hanging fruit and, to me, the lowest of that low hanging fruit is the security risk associated with using different relays.

http://www.eng-tips.com/viewthread.cfm?qid=409344

To be honest it is the first I am hearing this, however it makes sense when one thinks about it. I mean, how many substations have 3 transformers all of the same age and make to fulfill dependency/redundancy without issue? Now imagine all 3 being different, GE, ABB, Siemens. Now 3 different set of parts, manufactures and maintenance techniques all of which increase error. Perhaps where manufacturing defect is high such is a concern, but when all equipment is defect free, does it really matter? I could be wrong but just my way of thinking about... :?

 

[Bugman1400]

But have these issues not been ironed out? And since DF is relatively simple, no much can go wrong? My issue with 2 different relaying schemes from two different manufactures is a person has at least a 4x of misapplying a setting. At least that is how I see it IMHO.

Also, in regard to different schemes, dont such become harder to rely upon during N-1 or N-2 due to reduced fault current as you mentioned?

The issues that are known are typically ironed out but, its the unknown issues that are the ticking time bombs. However, I also totally agree with you about the misapplication of settings due to the unfamiliarity with different mfrs. If you were to weigh the risk of both, I think your second point would win.


The issue with DF (LD) is the high bandwidth that is needed. This relies heavily on a reliable comm..........which can be an oxymoron. It is not uncommon to have intermittent problems with comm during a thunderstorm when there is a high potential for faults. Sometimes, I see comm chattering for no reason at all. There is typically a team dedicated to chasing these elusive problems all over the system. For that reason, other schemes and protection layers need to be applied. You are correct that some schemes are more difficult to apply for n-1 or n-2 conditions which can affect the available fault current but, typically it affects the voltage source too so, that's why distance elements are a staple of protection.

 

[mbrooke]

The issues that are known are typically ironed out but, its the unknown issues that are the ticking time bombs. However, I also totally agree with you about the misapplication of settings due to the unfamiliarity with different mfrs. If you were to weigh the risk of both, I think your second point would win.


The issue with DF (LD) is the high bandwidth that is needed. This relies heavily on a reliable comm..........which can be an oxymoron. It is not uncommon to have intermittent problems with comm during a thunderstorm when there is a high potential for faults. Sometimes, I see comm chattering for no reason at all. There is typically a team dedicated to chasing these elusive problems all over the system. For that reason, other schemes and protection layers need to be applied. You are correct that some schemes are more difficult to apply for n-1 or n-2 conditions which can affect the available fault current but, typically it affects the voltage source too so, that's why distance elements are a staple of protection.

Can you elaborate more on voltage source in regards to relaying during n-1 and n-2? In regards to communication, is fiber optic (OPGW) effected by T storms?

 

[Bugman1400]

Can you elaborate more on voltage source in regards to relaying during n-1 and n-2? In regards to communication, is fiber optic (OPGW) effected by T storms?

The main issue with overcurrent protection are the huge variances with n-x conditions. Since distance protection uses V/I, the variation in current for n-x also affects the V part of the equation. So, a fault 3 miles out on a 20 mile line could have a fault current value of 5000 amps or 1000 amps depending on n-x. A typical overcurrent element would have trouble determining where the fault was on the line. However, the distance element is able to accurately measure that the fault is 3 miles out because as the bottom of the Z=V/I changes, the top V variable also changes to keep the impedance the same.

OPGW is not typically affected by T-storms, except from a direct hit where the static gets burnt. However, I typically equate T-storms with wind and that does affect OPGW at the splice boxes. Also, it is not uncommon for a fiber network to terminate into a uWave network. The uWave networks can have issues from wind, rain, lightning because the dishes get rocked.

 

[mbrooke]

Since a weak source would have more voltage dip and a strong source more voltage dip for any given fault current?