Line Diferential Relay l-90 doubt 2

[Malagaprot]

Hello to everyone. I,m responsible for de protecctíons in The Costa del Sol Spain. I recently put in service a L-90 from general Electric relay. When the line got in charge, I could see in the mettering inside the relay that the amperes in the local and in the remote relay where around 92A. The problem is that I had a differencial current from 13A. As I could see the fasors in the P.C I realiced that the problem was that there was a difference in angels of only around 8º (really 172º since the fasors are in oposition). I don,t know much about comunications but this department say that the comunications are ok. The T.I from one side are N>=15 in precition. I dont know if this corresponds to 10P20 or .... but the problem is not the presition but the angels. The problem with the angels is in every phase, so I can see that the 13A are on every phase.
Besides the T.I are diferent in both extremes but there is no problem with that since the current measured are the same 92A in every phase.

Can any one advise me of what to do?

Thanks in advance

Antxón

 

[DanDel]

I would first check the polarity of the CTs and their connections.

 

[tinfoil]

How long is this line?
At 50Hz, the wavelength of electrical propagation is 6000km.
8° (or 2.2% of 360°) represents about 133km of distance.
Your 8° phasor displacement might be 'real'.

 

[Malagaprot]

The polarity is correct. If not the angel would be 180º oposite and the diff current 180 instead of 13A.

The line length is 5.2km only.

There is something I did not mention, and that is that the T.i on one side are toroidal 800/5 5p30. and on the local 400/5 N>=15.

Thank you

 

[RalphChristie]

If you see a differential current in all your phases, your sets of CTs on the two sides are not the same. I have no idea what N>=15 means, maybe an CT-expert, like scottf, can help here.



From BS 3938:1973

For Protective CTs:
a 5P CT can have a phase angle error of ±60 minutes at rated primary current.
The phase angle error of a 10P CT is not specified.

For Metering CTs (maybe it is metering CTs)
Class 0.1
phase angle error between 10-20% of rated current = 10 minutes
phase angle error between 20-100% rated of current = 8 minutes
phase angle error between 100-120% rated of current = 5 minutes
Class 0.2
phase angle error between 10-20% of rated current = 20 minutes
phase angle error between 20-100% rated of current = 15 minutes
phase angle error between 100-120% rated of current = 10 minutes
Class 0.5
phase angle error between 10-20% of rated current = 60 minutes
phase angle error between 20-100% rated of current = 45 minutes
phase angle error between 100-120% rated of current = 30 minutes
Class 1
phase angle error between 10-20% of rated current = 120 minutes
phase angle error between 20-100% rated of current = 90 minutes
phase angle error between 100-120% rated of current = 60 minutes
Class 3 and Class 5
phase angle error not specified

The phase angle error is the phase displacement between the primary current and the secondary current of a CT, expressed in minutes. Besides the fact that you can have a current error between different CT-sets, you'll can also have a phase angle error between them.

Regards
Ralph

 

[RalphChristie]

Malagaprot

Also note:
The lower the %current through the CT in terms of rated current, the higher the possible phase angle error. A 5P30 800/5 CT can have a phase angle error of ±60 minutes at rated current, thus there can be a phase angle error of 1° (60 minutes) if 800A are flowing. At 80A (10% of rated current), the error will be much bigger.
The phase angle error of 10P CTs are not specified, resulting in bigger errors than 5P CTs.
An important considiration of feeder differential schemes are to ensure that all the CTs are the same.

Regards
Ralph

 

[alehman]

I would take a close look at the communication channel. You are told it is good. The relay requires a high quality channel with symmetrical propogation delay (unless other means such as GPS is used for synchronization). Digital channels may be asymmetrical while providing what would be considered by most communications folks to be a high quality link.

http://geindustrial.com/multilin/pr/l90cac.pdf

 

[mc5w]

The shunt capacitance to ground of a transmission line will make the supply end of the line to have a lightly more capacitive power for the load end at light load.

Under heavy load the opposite is true. In this case the series inductance of the line takes over and the supply end has a slightly more inductive power factor.

You could also have some leaks to ground in either your primary side wiring or your secondary side wiring. You could also have a water problem in your secondary wiring. Any secondary wiring that is below ground or otherwise subject to flooding of a raceway needs to be type XHHW or better yet the USE version of XHHW. XHHW insulation is more waterproof than THWN and XHHW has less capacitance especially when wet. I have had an instance where water in PVC conduit in the ceiling of a food plant had water in it which caused a variable frequency drive to not work. The wiring tested just fine at 60 Hertz and D.C. but at the carrier frequency was a poor excuse for a short circuit. The water acted as a giant capacitor plate that was in intimate contact with the insulation.

On the other hand, THWN is more fire resistant, more gasoline and oil resistant, and is more abrasion resistant. THWN also presents lower wire pulling friction as well.

 

[shirui]

I agree with alehman, 1ms asymmetry in communication will cause 1x360/50=18 degree difference in phase angle.