CT Voltage Class on low side of Gen Stator Windings

[proznick]

Hello all,

I am involved in design of 12.5MVA high resistance grounded (using NGT)hydro plant. Having a discussion on CT sizing on the stators low side. On the high side, bolted fault current is 8574A, which is mostly supplied by the utillity. CT's on the high side C800 600:5 rating.

My question is CT's on the low side of the generator. The low side stator conductors are only going to see generator supplied current and none from the utility. Which means the voltage class rating on these CT's can C100.

I can't see any fault situation where the low side of the stator conductor's, where CT2 is located, would be exposed to anything other than the generator supplied fault current.

I have attached a word doc showing different faults for this scenario.


Thanks in advance for any feedback on this.

RP

 

[davidbeach]

Are you thinking of applying differential protection on the generator?

If so you should match the performance characteristics of the CTs as closely as possible.

 

[proznick]

David,
Thanks for the reply.

Yes, using SEL300G for current differential protection on generator.

If CT's are operating in linear region of operation or within the 10% correction limitations for ultimate fault conditions, why is there a need to use the matched cores on the high and low sides of the stator as a design requirement. If they are Relay class CT's and operating within the bounds of saturation, the sensed current levels should be operating in a linear region for both the C100 600:5 on the low side of stator and the C800 600:5 on the high side of the stator and sensed currents should be accurate.

A C800 class relay is major overkill for the low side of the stator if it will only ever be exposed to the generators fault current levels.

I am fairly new to power engineering so appreciate you taking the time to respond, thanks.

RP

 

[slavag]

Hi Proroznic.
NO, NO and next time NO.
Please dont think about this way.
CT of the generator MUST be same, ALL of them.
not only class and burden, also package from factory.
It's a big mistake think about bolted fault only.
Problem is a autorecloseres, inrush of trafos and capacitoros and also not at the pich-up region..in time of current decay.
I see so much false trips of the generator diff 87G ptotection. I don't remember now, I think SEL have a good document on the issue.
Best Regards.
Slava.
Don't think about formulas, think about expereince
C100 is good maybe for the overcurrent protection, not more

 

[davidbeach]

At this point I'd pass the ball to your local SEL application engineer. For through faults, including faults on the high side of the neutral grounding transformer, both CTs must perform nearly identically, but they will perform differently. Maybe not differently enough to cause trouble, but may be. I wouldn't even think about mixing a C100 with a C800. If you can get away with it, you are spending way too much on the C800.

 

[scottf]

Proznick-

Just to be clear, you're not really asking about the "voltage class" of the CT, which normally means the insulation rating class, but rather the relay accuracy rating, in particular the accuracy limiting voltage rating, i.e. the "800" part of the C800 rating.

The reason it's recommended not to mix ratings in a differential protection scheme has to do with the difference in excitation currents in different rated CTs. Some of this concern dates back to electro-mechanical applications and often electronic relays aren't quite as sensitive to a mismatch in CT ratings.

With that said, from a pure risk/reward perspective, I can't imagine the savings in a C100 versus a C800 would come anywhere close to even bringing this to a decision point, given that it's a critical protection point/scheme.

 

[proznick]

Following DB's advice, I discussed this with an SEL applications engineer. As long as the homework is done in determining the fault currents and then sizing CT's appropriately to avoid saturation, you can use different core sized CT's in differential protection across the generator as desicribed in this scenario.

All your knowledgeable feedback appreciated, thanks

RP

 

[dpc]

While this may seem clever and it may work out using low burden digital relays, it still seems very false economy. There have situation where I was forced to use different accuracy class CTs on differential protection, but only because the CTs were already there.

One of the advantages of differential protection of machines as opposed to transformer differential relays is that the pickup can be set extremely low since there are no concerns about transformer taps and inrush currents. By going with C100 CTs, you may be giving up a lot of this sensitivity.

Since this is a new facility, you have a chance to do it right and not engineer in a compromise that someone will have to live with for the next 50 years just to save a few hundred dollars during construction. There is nothing in this plant more critical than the generator. This seems like a very odd place to be cutting corners.

 

[proznick]

DPC,
Point well taken, thanks

RP

 

[slavag]

Hi Proznic.
Please see attached thread on the issue.
thread238-209945
I woudn't and can't add something to posts of Scottf and Dave(dpc),I think it's more than clear.
Best Regards and Good Luck.
Slava