Exceeding Metering Class CT Burden Rating

[stuckbreaker]

Can the accuracy of a metering class CT be treated similarly to a C-rated protection class CT where the rating still holds as long as the voltage produced by the CT does not exceed its rating, or is the rated burden the maximum permissible connected burden?

For example, a C800 CT is rated at 20x nominal current (say 5A) and a burden of 8Ω. If it is connected to a load of 16ohms, the CT would hold its accuracy class at 10x nominal current (10*5A*16Ω = 800).

If I have a 100/5A 0.3B1.2 metering CT with TRF=4.0, the accuracy class should be held from 10% to 100%*TRF, so up to 20A. Does this mean that the connected burden can exceed the nominal burden rating as long as 20A*1.2Ω=24V is not exceeded? Would the metering CT still hold its accuracy at a connected burden of 2.4Ω as long as the current does not exceed 10A?

 

[System Protection]

In practice maybe. Technically I don't think there is a provision in C57.13 that requires accuracy above the standard burden though.

What does the CT manufacturer documentation say?

 

[bacon4life]

Extending the example to a rather extreme case, I suspect error would be quite large if you had a burden of 2400 Ω and current of 0.01 A.

 

[scottf]

Short answer: no

One of the main problems is that you're mixing transient performance of a protection class CT with "steady-state" performance of a metering class CT.

The accuracy of a metering class CT, is primarily a function of the % nominal current and the applied burden.

All else held equal, CT error gets "better" as the current gets higher (up to when it's close to saturation) and "worse" as the current gets lower.

All else held equal, CT error gets "better" as the applied burden is lower and "worse" as the applied burden is higher.

 

[stuckbreaker]

In practice maybe. Technically I don't think there is a provision in C57.13 that requires accuracy above the standard burden though.

What does the CT manufacturer documentation say?

I looked through C57.13 and couldn't find much. The datasheet I have doesn't say much - all I really have is the rating and TRF.

Extending the example to a rather extreme case, I suspect error would be quite large if you had a burden of 2400 Ω and current of 0.01 A.

I agree, but the accuracy rating wouldn't be guaranteed since the current is less than 10% of nominal. I suppose the extreme case would be a burden of 48 Ω and a current of 0.50 A.

Short answer: no

One of the main problems is that you're mixing transient performance of a protection class CT with "steady-state" performance of a metering class CT.

The accuracy of a metering class CT, is primarily a function of the % nominal current and the applied burden.

All else held equal, CT error gets "better" as the current gets higher (up to when it's close to saturation) and "worse" as the current gets lower.

All else held equal, CT error gets "better" as the applied burden is lower and "worse" as the applied burden is higher.

Is the C800 rating actually a transient performance rating? My understanding is that the C-class rating was more so related to the construction/size of the core. I would think that if you could somehow keep the CT/cables/equipment cooled and with sufficiently rated insulation, the C800 rating is valid continuously.

Can you provide any insight or literature on why the voltage argument doesn't work for a metering class CT? For a C-rated protection CT, it's pretty straightforward - the magnetic domains in the core are all aligned.