Smart questions
Smart answers
Smart people
Join Eng-Tips Forums

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

a10jp (Electrical) (OP)
30 Oct 09 7:59
I am trying to understand the method to detect ground fault if you have an ungrounded system.  In a few literatures I read that you will use relay 59G to sense gorunded overvoltage.  Because the ground fault current in this setup is very small, they said that you will require a sensitive device like directional overcurrent relay like device 67 in order to trip the circuit.  First, is this a common strategy for these type of system?  Is you have high resistance grounded circuit, would the situation be different?  When using HRG system, the benefit is the grounded voltage cna be stablized, as compared to unground sytsem, so, then, is that an issue with using device 67 which uses the grounded voltage (in ungrounded system) to determine a trip setting, knowing the setting range will be sensitive.  My apology if the question is confusing, as I am confused.
rbulsara (Electrical)
30 Oct 09 8:23
Which literature are you reading?

Perhaps you want to read some  in davidbeach's FAQ238-1287: What are good references for a Power Engineer? mentioned under Power System Protection.

Very briefly:

59G is most common for ungrounded and HRG systems. Since there is no current to speak of you depend on sensing the voltage distortion during a fault.

67 is just a directional device and has nothing to do with sensitivity. What you are referring to is perhaps the polarizing voltage. 67 do need a polarizing input (a reference for direction) either in form of a current or a voltage. The current sensing part still requires a CT input.

Sensitive ground fault detection is something different which can be applied to HRG system where current is very small but it still works the same a normal ground fault protection and needs an input from a GF CT. or residual connection.


Rafiq Bulsara

Helpful Member!(2)  slavag (Electrical)
30 Oct 09 10:09
Hi very shortly.
59G is goodest ground fault detection for whole type of grounding, but is protection per bus, per tansfromaer.

For ungrounded, HGR and compenstaed grunded systems used 67N function based on classical wattmetric or varmetrec methods.

You can use 51N too, but you need calculated possible capasitive current.

I recommend to you read great documents on the issue, from SEL, some registration, but isn't important.

Best Regards.
xxjohnh (Electrical)
30 Oct 09 10:17
In China they commonly install ungrounded MV systems. They have developed a special relays which is able to provide fault location. I think SEL has something similar that has been used as well. Contary to practice in EU and US  
slavag (Electrical)
31 Oct 09 0:52
Few contries are still used ungrounded MV systems.
In few places ground fault protection used for signal only, not for trip.
I don't want continieu with lot of problems of this networks, such of transient overvoltages, repetitive restrikes with less than 100ms, etc..etc... Big problem is cross-country faults. Up today used method of only two CT-s in such systems for avoid part of cross-countries faults and tripped two CB's instead one.

Protection methods for such systems are:
1. Varmetric method : U0xI0xsinPfi(0) or I0xsinPhi(0) and U0 as starter.
2. Method are base on high harmonics in the strikes.
3. Injection method.

Lot of companies have a good products with years of expirience: ABB, VAMP, Siemens.

Siemens, for exapmle, used letters 67Ns, VAMP CAP/RES, ABB cos/sin Pfi, for protections in ungrounded and compenesated systems.

Don't forget, you have ground fault in ungrounded systems too, it's capacitive current, for 20kV cable is about 4-5A per 1km.

Best Regards.
a10jp (Electrical) (OP)
31 Oct 09 4:05
Thank you.  Just want to add clarify the picture a little more.  The system we have is 13.2KV ungrounded system.  I did not underatand why we do it this way, but that is what we have.  Method of protection was to use 59G to detect, and use 67 to trip.  Earlier, rbulsara mentioned ungrounded and HRG system worked the same way as normal ground fault protection and utilize the GF CT for input.  When I0 is only a few A, can we still use CTs for such small current?  Now I am wondering if device 64 can be used.  Also, the system I am looking at use only 67, but not 67N.  What is the main difference between 67 and 67N that have been mentioned earlier?
slavag (Electrical)
31 Oct 09 4:45
67- is directional overcurrent protection
67N- is directional earth/ground protection
67Ns- is directional sensetive earth/ground protection.
See here possible some porblems

67N is directional, base on the I0, U0 and angle between them. Isn't exactly good for ungrounded system.
Siemens use 67Ns naming.
SEL, if I remeber right, used 32W or 32V naming.

Size of depend. I hev expirence with ungrounded system only once. 6.9kV.
We install toroid 100/1A ( now burnt2x saied something smile )
connect 1A to 0.2A input of relay and put minimum setting 1%. That means 0.2A primary.
It's work w/o any problem.
Always Q, why 100/1A, this size have a goodest performence for angle and amplitude errors for the small currents.

Hope that help.
Best Regards.
odlanor (Electrical)
1 Nov 09 5:59
There is a very good explanation about this problem in this Sell article:
Advanced Commercial Power System Protection Practices Applied to Naval Medium Voltage Power Systems.
Helpful Member!(2)  prc (Electrical)
1 Nov 09 22:33
slavag,The system I had seen for 11 &33 kV ungrounded system was the use of three single pahse PTs or one 5 limbed PT connected to lines with one of the PT secondaries in broken delta to sense voltage during ground fault.How this compares with the practices mentioned by you.
slavag (Electrical)
2 Nov 09 0:33
Sorry prc.
I don't know about this method.
Are you have some scheme of such connection? Could you  please attach it and we'll try understand, what is it.

Best Regards.

prc (Electrical)
2 Nov 09 5:53
slavag,We use such a scheme to find out inadvertant earthing of tertiary bus of auto transformers. Single phase Pts will be connected to line with secondary  windings of 110/3 V winding will be connected in broken delta.The ends of broken delta will be connected to a voltage sensing relay.Normally voltage across the corners ie ends of broken delta will be zero.In case of a solid ground fault on any line,voltage across the secondary broken delta will rise to 110 V raising alarm or trip.
slavag (Electrical)
2 Nov 09 11:57
Hi prc.
OK, now is clear.
It's classical connection of broken delta for tertiary AT winding ( delta)protection. We used it too, btw, tertiary AT(T) winding in our area is not used.
It's simple 59N function, not directional, as I was posted
"59G ( 59N) is goodest ground fault detection for whole type of grounding, but is protection per bus, per transformer".

Best Regards.
a10jp (Electrical) (OP)
2 Nov 09 21:47
Hi Slava, how are 59N and 59G different?
slavag (Electrical)
3 Nov 09 0:03
It's same-residual overvoltage protection, connected to broken delta of VT secondary..
Best Regards.
a10jp (Electrical) (OP)
3 Nov 09 8:17
Hi, the system we have used a parameter called neutral displacement which utilizes 3*V0 as the input.  The pickup level is set at 0.85 *VT using a definite time curve as part of the overvoltage detection.  Our first mistake was to set this all in trip mode.  But now we learned you cannot properly trip a circuit (using only V0) until you know where the fault is, which is where 67 is used for directional I0 element.  Make sense?
slavag (Electrical)
3 Nov 09 10:08
Yep it's possible too.
BTW, neutral displacment and residual voltage and  zero seq. voltage are same. Neutral displacment is more Siemens's naming.

And not 67 element, 67N element ( if it Siemens relays 67Ns function). You need take in account biggedt problem of such system, cross country faults.
For reduce percentage of such faults, used only two CT's and toroidal CT on all 3-ph. 0.85 seems a big value. Are you use broken delta VT connection or internal calculated?

Best Regards.
Bronzeado (Electrical)
3 Nov 09 11:36
Hi folks,

I wonder if the phase-to-earth voltage in an isolated (ungrounded) system can reach values more than phase-to-phase voltage.

Best regards,

Herivelto Bronzeado
odlanor (Electrical)
3 Nov 09 11:39
We have been experienced with these protections. Never use 67 to trip!It is too much sensitive and you could have many maloperation!
It is used in power plant with multiple generator paralleled at same busbar and connected  one step up transformer. Generating Voltage system is high impedance grounding by each neutral of generator(5A).
Each generator has 59N and 67NS very sensitive connected   at each CT summation close to busbar.
There are 2 trips:
67NS + 59N => trip 52G
59N with delay => trip 52G

Helpful Member!  davidbeach (Electrical)
3 Nov 09 13:28
Bronzeado, yes.  Arcing ground faults can develop phase-ground voltages well in excess of phase-phase voltage; voltage builds up until something flashes over.
slavag (Electrical)
3 Nov 09 14:26
In additional to David.
Especialy in case of time betweem restrikes is less than 100ms.
Bronzeado (Electrical)
3 Nov 09 14:38
Thank you davidbeach&slavag.

Do you have literature on this phenomenon (neutral instability)?


davidbeach (Electrical)
3 Nov 09 15:22
See FAQ238-1287: What are good references for a Power Engineer?.  In particular look to Blackburn's Protection book.
Bronzeado (Electrical)
3 Nov 09 16:59
Thanks you, davidbeach.
A star for you!


a10jp (Electrical) (OP)
3 Nov 09 20:34
67N and 67G are probably the same thing?
prc (Electrical)
3 Nov 09 23:16
it is almost the same.To decipher the numerous numbers and letters that  protection experts use,refer to IEEE std C37.2-2008 Device Function Numbers,acronyms  and contact designation.Equivalent IEC is 61850-7-4.

"The suffix N is preferred when the device is connected in the residual of a polyphase circuit, is connected across a broken delta, or is internally derived from the polyphase current or voltage quantities. The suffix G is preferred where the measured quantity is in the path to ground, or, in the case of ground fault detectors, is the current flowing to ground. See Figure C.2 in Annex C for examples."

Similarly "Device number 59—overvoltage relay
A device that operates when its input voltage exceeds a predetermined value." etc,etc

Let me share some "stories" on the subject.

In mid 70's, when one of the earliest 400 KV transformer banks were commisioned in India-it was a 250 MVA bank with 11 kV tertiary formed by power cables,one corner of the delta was earthed.After a couple of monsoon rains,one of the cable developed ground fault creating a L-L fault on tertiary causing tertiary winding failure.

To overcome this, we removed one corner earthing and put three single phase PTs (with primary neutral grounded) at tertiary terminals with 110/3 V secondary winding connected in broken delta to 59N relay.So in case of a LG fault on any line,59N relay will pick up 110 V residual voltage to give alarm.

After some years,during mid 80's I had to face  a case where 20-50 v was coming across the broken delta without any apparant earthing on tertiary bus anywhere. Cliet was not ready to accept the transformer suspecting some defect.
Then I found out that American GE and Westinghouse engineers had experienced the same problem in ungrounded MV systems during mid  40's and they explained the reason for this neutral displacement as the resonance between the inductance and capacitance of the tertiary circuit.Solution for this is to connect surge absorbers (capacitors) to tertiary terminals (which we did) or connect a resistor in the closed delta of the PT.Those old AIEE papers saved my skin at that time.
a10jp (Electrical) (OP)
3 Nov 09 23:56
Thanks for your help prc.  I am reading a bit on this subject lately on ungrounded system.  If you can point me to some of these earlier references, i would love to read about them.  

I, too, obserevd that all the feeder circuit in our system has a surge protection/arrestor device on EVERY feeder circuit, I believe it conforms to your statement before.

Also, after readin all the discussion above and tech papers in the referneces, at this point, I wanted to ask my original question:  is the method that we have discussed so far a common method used for ungrounded system, with applications between 59N and 67N?  Are there other methods available?
slavag (Electrical)
4 Nov 09 0:18
Thanks prc.

Answer-yes. I would like say again, please check what type of 67N you will be use. I strongly recommned use I0xsin(phi) operation criteria. Is good and work w/o problem. Maybe, here, we have guys from Finland, them have a good expirience with such systems.
59N you can use for alarm and/or back up trip.
67N/67Ns/34Q for detect, what feeder is faulty.

BTW, in newer realys, you don't need additional 59N relay for start 67N function, you have parameter U0 in relay.

That menas, criteria will be I0xsin(phi) and U0.

Best Regards.
prc (Electrical)
4 Nov 09 0:33
a10jp, I had the  list of all those seminal AIEE papers which Iam not able to locate now.Those were really wonderful papers going deep in to fundamentals.

I suggest the following papers which you can find in the respective sites.

1) A review of system grounding methods- Gerald Johnson

2)Ground fault protection on ungrounded systems-Application


Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Back To Forum

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close