Relay setting for dummies

[benbo]

I originally posted this elsewhere, but figured it might get more attention here.

Speaking of relays - here is a question for power experts. As you may or may not know, I am about as far from a relay expert as it gets. Is there some rule for where you should set the minimum pickup on a type 51 relay for a high voltage line? Is there a damage curve that you look at? Is there any relation to the ampacity? I read somewhere that for MV it is recommended to set it at or below the ampacity, but sometimes you can set it 300% above. But maybe that was if you also have a fuse. Assume the only thing I'm worried about protecting is an ACSR conductor.

I googled and can't seem to get a definitive answer. I have the NEC, but don't think that applies to this type of line - and i am terrible at using that anyway. I don't have the NESC, and not sure this type of thing would even be in there.

And don't worry. Nobody is hiring me to do a coordination plan.

I just want to know if there is a rule of thumb.

 

[HornTootinEE]

NEC and NESC Do not apply to an HV line owned by the utility.

I work in distribution, and I set forward overcurrent (51/51N) elements somewhere between 150% and 200% of full load, depending on the particular recloser, transformer high side fuse curve, downstream fusing, underground that may have lower fault withstand, whatever. Negative sequence elements may or may not following the same philosophy.

System protection guys are pretty diverse as to their opinions and rule of thumbs. Hence why Mason's book was called "The Art and Science of Protective Relaying" There is an art to it. I worked in transmission/substation system protection for my first year out of school. I learned a ton, and some was that protection philoshy changes with the guy behind the wheel. The person who mentored me that year is now retired and his understudy that replaced him has different setting philosophies in a lot of ways.

 

[cableguy]

I agree, relay setting is all about philosophy. The head of our department has been doing it here since the mid 1970s. So there's definitely an ingrained philosophy that our company takes towards settings... that may or may not be right for other companies. We have some "cookbook" settings sheets that we work from, but there are exceptions to every rule. And just when you think you get it, along comes a curveball...

I told our manager that I was still working on my requisition for the crystal ball, so that I can get more of these settings.

The trouble is not the individual settings... 51 by itself is fine. But when you use a Schweitzer 421 relay, which has over a dozen different devices built in to it, with hundreds of settings, knowing which ones to use and when is the big trick. That's where I get confusified. And my company seems to believe in OJT by osmosis... no classes or anything formal.

 

[Mary Faye]

The trouble is not the individual settings... 51 by itself is fine. But when you use a Schweitzer 421 relay, which has over a dozen different devices built in to it, with hundreds of settings, knowing which ones to use and when is the big trick. That's where I get confusified. And my company seems to believe in OJT by osmosis... no classes or anything formal.

I know where you are coming from. My company has the same view of training. One thing I have learned is that SEL engineers are very helpful and very willing to help. Sometimes they even help you ask the question

 

[cableguy]

Yeah, I talk to my local applications engineer several times a week, lol. I think I got some folks at SEL pretty riled up about their 5045 (TEAM) software ($$$$$). We bought it, and it's one of my pet projects. It doesn't work properly for us (yet). Keep finding major bugs. They promise a lot, but so far, all I have is software that may run for a day or two at most before it's hosed. It has potential, but it's definitely in the "not ready for prime time" category.

SEL makes fantastic relays, but much of their Windows software is sub par. I base that on my experience with 5010, 5030, 5040, and now 5045.

I'll eat my crow when SEL makes good on the software (I hope, I hope). But they've got a long way to go. I'm doing my best to help them improve their software though. It's pretty sad when in one day I can write a few lines of php and come up with a web interface to display information in their database that they do not show on any screens (stuff that I consider non-trivial, like communications statistics - I want to know how my system is performing, the info is in the database, yet it's not available on any SEL displays). A relay engineer might not be as interested in comm stats, but having been a SCADA/comms guy for 15+ years (and just switching over to being a relay engineer), I am most definitely interested.

 

[HornTootinEE]

I have found the SEL field engineers to be very very helpful. And if you're stuck, our guy out here will tell you give him a couple days, and he'll get back to you with a full out settings solution for what you are trying to do. Our relay guru in-house used to do that for us, but he is the guy that retired, so learning curves are accelerating around our company.

We have the same training philosophy, figure it out or die trying.

 

[Wolverine]

Aha, you have asked a very complicated question that does not have such a simple answer.

Actually, when setting a relay, you're more concerned with fault current, not line loading. I had it burned into me early on that WE DO NOT SET RELAYS FOR LOAD! That being said, you want to be at least somewhere above 120% to 140% of full load (the more precise the relay, the lower you can go). More importantly though, you want the relays set as low as possible while still maintaining consecutive device coordination. What that means is that you want your relay (& fuse) curves to coordinate at the lowest level possible (yes there is a curve that defines how fast a 51 trips based on fault current, and just to make it more complicated, there are many different curves to choose from - slow, fast, very fast, ... even shaped curves in the smarter relays).

At the end of the day you want your high side transformer relay to coordinate with low side, your low side transformer to coordinate with your line relay, your line relay to coordinate above your load fuse, and your fuse set so that it trips at 100.1% of line capacity, but only after a long time delay, and fast enough to fit under any equipment damage curves.

In ascii characters, it looks something like this: (((((

The more fault current, the faster the trip.

Every operating company has it's own rules of thumb. Ours feeder settings were based on source transformer size. It gets to be even more fun when you realize an elecro-mechanical relay has fixed taps, so you either have set it to 5X or 6X, not between (unless you know how to bend the spring). Smart relays can be set to the decimal place.

All this is assuming you are talking about radial lines. Network lines are a whole different story. 'Tis truly an art.

 

[benbo]

What that means is that you want your relay (& fuse) curves to coordinate at the lowest level possible (yes there is a curve that defines how fast a 51 trips based on fault current, and just to make it more complicated, there are many different curves to choose from - slow, fast, very fast, ... even shaped curves in the smarter relays).

Yeah, actually I know a little about how the relay operates, with very inverse, extremely inverse, etc. So I basically know how they function once they are set, but not where they should be set.

At the end of the day you want your high side transformer relay to coordinate with low side, your low side transformer to coordinate with your line relay, your line relay to coordinate above your load fuse, and your fuse set so that it trips at 100.1% of line capacity, but only after a long time delay, and fast enough to fit under any equipment damage curves.

So I guess my basic question would be this - If the relay operates as it was supposed to, picking up at the right fault current, and tripping after the set delay time, but the line burns up anyway, would it be correct to say there is probably something wrong with the way some part of the protection was set up? Or do you look to potentially (under some rare circumstance) sacrifice some part of the system to protect more expensive or critical parts? I'm assuming you don't want these things tripping all the time when they shouldn't, but I also assumed that's what reclosers were used for.

 

[HornTootinEE]

There is the art. Coordination while preventing unwanted trips and protecting equipment at a reasonable level to prevent damage due to through faults.

 

[benbo]

Okay, thanks guys. That helps a lot. THese two things help a lot -

At the end of the day you want your high side transformer relay to coordinate with low side, your low side transformer to coordinate with your line relay, your line relay to coordinate above your load fuse, and your fuse set so that it trips at 100.1% of line capacity, but only after a long time delay, and fast enough to fit under any equipment damage curves.

There is the art. Coordination while preventing unwanted trips and protecting equipment at a reasonable level to prevent damage due to through faults.

I can see there's not necessrily a set answer. It's sort of like whack a mole.