The advantage of a residual overcurrent relay vs a ground fault relay 1

[majesus]

I'm curious when you would favor using a residual short circuit relay for gnd fault detection (relay 50N) versus a ground fault relay (50GS) and (vice versa.)

I came up with (these are just a guess):
1) 50N requires only one CT on all conductors thus can detect phase faults as well.

2) A 50GS is possibly easier and cheaper to install if the grounding conductor is available from Neutral to the grounded Electrode.

3) A 50GS has better accuracy?

 

[davidbeach]

Don't know that I've ever seen the notation 50GS. Several of the terms you used can be ambiguous, so why don't you provide a more detailed description of the methods you are asking about? Some of your description and some of the terms used aren't the most common combinations of those concepts, at least in my experience. It'll be easier to answer the questions if I (or others) have a better understanding of the question.

 

[lansford]

In simple terms, a residually connected ground relay is more appropriate for a solidly, multi grounded wye system.
If the system is resistance grounded in the range of 25 to 2000 amps, then a zero sequence (window CT) feeding a ground relay is more appropriate.
JIM

 

[majesus]

Neither did I Davidbeach...

So here is what I mean:
This is the SLD where I saw this:

http://i48.photobucket.com/albums/f223/majesus/Eng stuff/50GSSLD.jpg

I did a google on it and came up with this document here from GE called “Relay Selection Guide”:
On Page 21 of the PDF it says:
"51N Residual Overcurrent"
"50GS Instantaneous Ground Fault Relay"
So I started to think on the question: "In application, when to choose one over the other?"

My list of reasoning in my first post was just me brain storming on the possibilities... which sometimes is far from the truth!

Lansford, I thought a residually connected ground CT is a window CT? (Not the other way around, ie zero sequence -> Window CT)
The reason I say this is because in the GE pdf it states a 51N as Residual and the 50GS as Instantaneous. However it shows that the 50GS having a dedicated ground relay feeding the relay as shown in the figure on page 21. The 50N does not (refer to figure on page 20 of the manual).

My notes from class show a window type as one relay over all phases (the 50N?) and the grounding relay in Figure B (the 50GS? Just like as shown in the GE pdf on page 21)

http://i48.photobucket.com/albums/f223/majesus/Eng stuff/notes.jpg

And as always, I appreciate the input

 

[majesus]

Sorry, I forgot to post the link for the GE pdf document called: called “Relay Selection Guide”

http://www.geindustrial.com/multilin/notes/GET-8048A.pdf

 

[dpc]

In my experience, "50GS" has always been used for a ground relay with a dedicated zero sequence (flux summation) CT/sensor. "Residual" has always meant a relay or relay element connected to the return conductor of three phase CTs used for phase overcurrent. It is measuring the "residual" current in the CT secondary circuit - the sum of the three phase currents.

 

[DanDel]

'Residual' typically refers to a connection in the summation of the phase and neutral CT secondary wires, as dpc said. This is usually refered to as a 50G device.

'Zero Sequence' (or core balance) is used for a relay connected to the secondary of a single large window CT over the three phase conductors or the phase and neutral conductors.

50N is typically used for a GF protection scheme that uses a single CT in the ground return connection of the transformer.

I have seen instances where the 50G and 50N designations have been used interchangeably, though.

 

[majesus]

Thanks guys...

I have made up a schematic of the three:
50N, 50G and 50GS. Is it correct?

http://i48.photobucket.com/albums/f223/majesus/Eng stuff/Relay3.jpg

Maj

 

[davidbeach]

What about calculated residual current when the relay gets the three phase currents and calculates the zero-sequence current?

 

[majesus]

Are you talking about my Relay Labelled "R" with an arrow pointing as "50N Relay"?
It trips the same ground fault sensor as pointed to by the 50GS Relay...

I didn't draw it in as it would an extra line going across the page... I just want to be sure I know the difference types of Ground Sensing Relay schematics and there labeling (50N, 50G and 50GS)

 

[majesus]

I'm sorry, I don't want to repost my same question, but I just want to know if I labeled these type of relays correctly. Confirmation as "yes those relays are a 50N, 50GS and 50G as drawn." Or "NO you are completely totally lost."

Here is the link again for convenience:

http://i48.photobucket.com/albums/f223/majesus/Eng stuff/Relay3.jpg

 

[davidbeach]

There are various ways of labeling it. I work in the higher voltages, where some of those schemes just don't work.

A 50N or 50/51N could be in the residual connection, as you call a 50N; or it could be on the neutral-ground connection of a transformer or generator. With the relays I generally work with, a 50G or 50/51G is calculated residual inside a numeric relay as opposed to measuring the current in through the IN current terminals.

 

[JensenDrive]

1) In you figure, I think what you referred to as 50G is an uncommon device, but if it exists, I suppose 50G would be a correct name. However, this 50G would only work at one point in a typical low voltage (LV) power system. Neutral is only connected to ground at one point in an LV system (at the service entrance). I do not know how often people put overcurrent elements on the tie between neutral and ground, but I have a suspicion it is not normally seen. Hopefully, others will comment.

2) Low voltage breakers tend to have a separate CT around the neutral, and the output of that CT is added to the phase CTs inside the trip unit, so your 50N is closest to what goes on in LV breakers. However, I do not recall seeing the term 50x used when referring to the standard ground fault element in a low voltage circuit breaker. It is just called the ground fault element. Hence, your 50N is not a term used in the application you show, but it exists functionally in LV circuit breakers.

3) In some LV breaker applications, people want an extra sensitive ground fault element, so they buy LV breakers without the standard ground fault element and install a separate ground fault element that monitors a CT that wraps all 3 phases + the neutral. This is what you show as 50GS. I read the "GS" as "Ground-Sensitive" and I think that it is a satisfactory abbreviation.

4) I prefer to say "3Io" when referring to the calculated residual current (Ia+Ib+Ic). This in turn means that in a system that carries a separate neutral that "ground" is Ia+Ib+Ic+In OR the output of a CT that wraps all 3 phase + neutral.

5) In medium voltage systems and relays, there is no true identifiable neutral current except at the neutral of xfmrs and generators, and the use of G vs N for ground is all over the map; they are almost interchangeably. You have to take it in context. However, I think the preponderance of cases tend to use N when referring to a) 3Io current or b) current in the neutral of a generator or xfmr, and G to refer to the current actually flowing in the ground.

6) Manufacturers of relays are not consistent on the use of N and G. The SEL321 and their earlier 121x and 221x relays used "N" to refer to 50 and 51 elements that respond to 3Io, but the SEL311x series they changed their mind and decided to use 50G and 51G for the same current. But even within the 311x relays, they label some of their 3Io fault detectors as 50Gx elements. I have seen the same inconsistency in other mfrs relay lines. Some relays are on the market that have the 3Io calculation and also have a separate 4th CT input, and I think they tend to use In for 3Io, and Ig for the 4th CT input. However, some mfrs refer to the 4th CT input as the "sensitive earth" current.

 

[majesus]

Beautiful JensenDrive. Thanks for the explanation, makes sense. Very very useful. Especially the knowledge about defining relay types used in LV and MV.

Saved in my reference directory. In the near future, I'm sure I will come across this again (time and time again.)

BTW, I got the info to make up my drawings from here:

http://polk-burnett.apogee.net/pd/dvprgo.asp

 

[jghrist]

majesus,

Note that your Polk-Burnett reference uses 50G differently than you did.

Neutral ground relaying uses a time-delay overcurrent relay (51G) connected in the secondary of the current transformer located in the neutral of a wye-connected transformer, wye-connected generator, or the neutral of a neutral-deriving transformer bank.

This is the usage referenced by JensenDrive's 5)b).